Azetidine and Piperidine Carbamates as Efficient, Covalent Inhibitors of Monoacylglycerol Lipase

Article abstract of DOI:10.1021/acs.jmedchem.7b01531

Monoacylglycerol lipase (MAGL) is the main enzyme responsible for degradation of the endocannabinoid 2-arachidonoylglycerol (2-AG) in the CNS. MAGL catalyzes the conversion of 2-AG to arachidonic acid (AA), a precursor to the proinflammatory eicosannoids such as prostaglandins. Herein we describe highly efficient MAGL inhibitors, identified through a parallel medicinal chemistry approach that highlighted the improved efficiency of azetidine and piperidine-derived carbamates. The discovery and optimization of 3-substituted azetidine carbamate irreversible inhibitors of MAGL were aided by the generation of inhibitor-bound MAGL crystal structures. Compound 6, a highly efficient and selective MAGL inhibitor against recombinant enzyme and in a cellular context, was tested in vivo and shown to elevate central 2-AG levels at a 10 mg/kg dose.

Full text of DOI:10.1021/acs.jmedchem.7b01531

Subscriber access provided by READING UNIV
Article
Azetidine and Piperidine Carbamates as Efficient,
Covalent Inhibitors of Monoacylglycerol Lipase
Christopher R. Butler, Elizabeth M. Beck, Anthony R. Harris, Zhen Huang, Laura A. McAllister, Christopher
W. am Ende, Kimberly F. Fennell, Timothy L. Foley, Kari R. Fonseca, Steven J. Hawrylik, Douglas S.
Johnson, John D. Knafels, Scot Mente, Stephen Noell, Jayvardhan Pandit, Tracy B. Phillips, Justin
R Piro, Bruce N. Rogers, Tarek A. Samad, Jane Wang, Shuangyi Wan, and Michael A. Brodney
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01531 • Publication Date (Web): 17 Nov 2017
Downloaded from http://pubs.acs.org on November 17, 2017
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Azetidine and Piperidine Carbamates as Efficient,
Covalent Inhibitors of Monoacylglycerol Lipase
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
Christopher R. Butler*a, Elizabeth M. Becka, Anthony Harrisb, Zhen Huanga, Laura A. McAllistera,
Christopher W. am Endeb, Kimberly Fennellb, Timothy L. Foleyb, Kari Fonsecaa, Steven J. Hawrylikb,
Douglas S. Johnsona, John D. Knafelsb, Scot Mentea, G. Stephen Noellb, Jayvardhan Panditb, Tracy B.
Phillipsb, Justin R. Piroa, Bruce N. Rogersb, Tarek A. Samada, Jane Wangc, Shuangyi Wanc, and Michael
A. Brodneya
b
aPfizer Worldwide Research and Development, 1 Portland St, Cambridge, MA, 02139, United States;
Pfizer Worldwide Research and Development, 445 Eastern Point Road, Groton, CT, 06340, United
States,cWuXi AppTec 288 Fute Zhong Road Waigaoqiao Free Trade Zone Shanghai 200131 China
RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required
ABSTRACT
Monoacylglycerol lipase (MAGL) is the main enzyme responsible for degradation of the
endocannabinoid 2ꢀarachidonoylglycerol (2ꢀAG) in the CNS. MAGL catalyzes the conversion of 2ꢀAG
to arachidonic acid (AA), a precursor to the proꢀinflammatory eicosannoids such as prostaglandins.
Herein we describe highly efficient MAGL inhibitors, identified through a parallel medicinal chemistry
approach that highlighted the improved efficiency of azetidine and piperidineꢀderived carbamates. The
discovery and optimization of 3ꢀsubstituted azetidine carbamate irreversible inhibitors of MAGL was
aided by the generation of inhibitorꢀbound MAGL crystal structures. Compound 6, a highly efficient and
selective MAGL inhibitor against both recombinant enzyme and in a cellular context, was tested in vivo
and shown to elevate central 2ꢀAG levels at a 10 mg/kg dose.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 48
1
2
3
4
INTRODUCTION
5
6
7
8
9
Inflammation serves as an evolutionarily conserved protective mechanism to reduce infections
and clear antigens as well as cellular debris to minimize tissue damage and promote subsequent tissue
healing following the resolution of the inflammatory cascade. Exaggerated or persistent inflammatory
response in the periphery or the central nervous system (CNS) can become pathological, leading to
chronic inflammatory conditions. Neuronal homeostasis in the CNS is particularly vulnerable to chronic
inflammation, and most neurological disorders are associated with an imbalanced inflammatory response
in the brain. Hallmarks of neuroinflammation include activation of resident immune cells (microglia and
astrocytes), production of proꢀinflammatory cytokines and chemokines, as well as deterioration of the
bloodꢀbrain barrier. This cascade leads to heightened risk of neuronal dysfunction and neurodegeneration
and in some cases infiltration of peripheral cells and proteins into the CNS.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
We have previously shown that monoacylglycerol lipase (MAGL), an enzyme that terminates the
signaling of the endocannabinoid 2ꢀarachidonoylglycerol (2ꢀAG), also controls arachidonic acid (AA)
production in the brain.2ꢀ3 MAGL catalyzes the hydrolytic conversion of 2ꢀAG to arachidonic acid and
glycerol.2 2ꢀAG accounts for 50% of the production of AA in the brain,4ꢀ5 whereas phospholipid
hydrolysis by phospholipase A2 (PLA2) serves as the primary source of arachidonic acid production in
the periphery. AA plays a major role in the inflammatory response, both as a signaling molecule, but
also as a precursor to multiple pathways of proꢀinflammatory eicosanoids (Figure 1). As such, MAGL
exerts a bidirectional control on brain inflammation by modulating brain endocannabinoid and
arachidonate levels in opposing directions. Therefore, MAGL inhibition potently decreases the proꢀ
inflammatory eicosanoids and cytokines in the brain5 and enhances 2ꢀAG signaling at CB1/2 receptors.
Genetic deletion or pharmacological blockade of MAGL attenuates biomarkers of neuroinflammation, 6ꢀ11
including cytokines and gliosis, in mouse models of Alzheimer’s disease.12 The use of steroids is
currently the main therapeutic option to attenuate pronounced brain inflammation, therefore development
ACS Paragon Plus Environment

Page 3 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
of potent, selective inhibitors of MAGL represents a research avenue with significant therapeutic potential
for the treatment of acute and chronic diseases associated with neuroinflammation.
7
8
9
Figure 1. Role of MAGL in Endocannabinoid and Arachidonic Acid pathways
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
MAGL is a 33 kDa serine hydrolase with an active site containing a classical SerꢀHisꢀAsp
catalytic triad.13 The characteristic reactivity of the active site enables the opportunity for covalent
modification of the key serine residue (Ser 122). Covalent inhibition is attractive in that it offers the
potential for extended duration of pharmacodynamic modulation relative to pharmacokinetic profile of the
inhibitor.14 Thus, a sustained elevated pharmacological response can be achieved without the
corresponding sustained exposure, as would be required for a reversible, nonꢀcovalent inhibitor. In order
to take advantage of these potential attributes, covalent inhibitors require two key elements: a) sufficient
reactivity for the residue that is to be modified in the protein and b) specific target binding affinity.
Serine hydrolases as a class have been successfully inhibited by a number of electrophileꢀcontaining
molecules that range in their inherent reactivity.15ꢀ16 More specifically, inhibition of MAGL has been
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 48
1
2
3
4
5
6
7
8
9
accomplished most often in the literature with a carbamate electrophile that efficiently engages Ser 122,
as observed in JZLꢀ184 (1, Figure 2)3, 17 but is not susceptible to subsequent catalytic turnover, thus
inhibiting the enzyme. In order to facilitate inhibitory acyl transfer with Ser 122, these carbamates
generally contain an alcoholic subunit with a relatively low pKa to serve as a suitable leaving group. As
shown in Figure 2, exemplar literature MAGL inhibitors typically employ alcohols such as 4ꢀnitrophenol
and hexafluoroisopropyl (HFIP) alcohol,.18ꢀ21
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
NO2
O
O
O
CF3
CF3
O
N
O
OH
N
O
H
O
N
S
O
O
1
2
O
Cl
MAGL IC50 = 5.7 nM
FAAH IC50 > 11 M
MAGL IC50 = 1.4 nM
FAAH IC50 = 0.80
M
Figure 2. Literature MAGL inhibitors. Data shown obtained from inꢀhouse assays
In addition to the requirement of a suitable leaving group, two key structural elements for the
inhibition of MAGL can be elucidated in the comparison of the structural features of 2ꢀAG to known
inhibitors.13 As illustrated by both the lipophilic arachidonate and bisꢀdioxolane tail groups in 1,
significant lipophilicity is required distal to the catalytic serine (Figure 3, pink). This lipophilic tail is
attached via a linear linker (yellow) that partially recapitulates the alkyl chain of 2ꢀAG. These structural
requirements underpin the challenge in developing orally bioavailable MAGL inhibitors suitable for inꢀ
vivo evaluation: incorporating the requisite lipophilic tail region and leaving group (blue) without
dramatically increasing clearance, which is known to be unfavorably impacted by increased
lipophilicity.22 Early optimization of the linker becomes crucial to facilitate subsequent tuning of
lipophilicity and molecular weight in the tail region. This letter will describe the identification of
efficient core amine systems by monitoring the optimization using Lipophilic Efficiency (LipE) and Fit
ACS Paragon Plus Environment

Page 5 of 48
Journal of Medicinal Chemistry
1
2
3
4
Quality (FQ).23ꢀ24
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
Figure 3. Arachidonic Acid and related pharmacophore
Results and Discussion
With the goal of identifying a potent, brainꢀpenetrant, and low clearance MAGL inhibitor, initial
efforts for linker optimization occurred via two related approaches enabled by the diversity available from
Pfizer’s molecule collection (Figure 4). The first approach (A) was to utilize small, fragmentꢀlike
monofunctional amines, lacking the lipophilic tail portion of the molecule, to prepare HFIP carbamates.
The second approach (B) utilized biꢀfunctional monomers, either by incorporating an equivalent
lipophilic tail, deprotecting, and then forming the key HFIP carbamate, or alternatively, inverting the
sequence to introduce the carbamate, and subsequent deprotection and incorporation of lipophilicity.
Approach A: Carbamate formation directly from monofunctional amines
Approach B: Diamine functionalization and carbamate functionalization
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 48
1
2
3
4
5
6
7
8
1) TFA, CH2Cl2
N
N
N
R
O
N
Boc
2) triphosgene,
Et3N, HFIP
DMAP, CH2Cl2
R
R
F3C
O
R = CHO or CH2Br
CF3
9
NH
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
N
Boc
R
O
CF3
CF3
O
CF3
CF3
1) TFA, CH2Cl2
2)
triphosgene,
Et3N, HFIP
DMAP, CH2Cl2
N
O
N
O
N
N
R
Boc
R
R
R = CHO or CH2Br
Figure 4: Parallel approaches to linker exploration
Whereas the combination of these two approaches significantly broadened the availability of
linker architecture, it required a normalization of potency to account for the stark differences in
lipophilicity and molecular weight between the two approaches, and was ultimately accomplished using
LipE and FQ. LipE, (calculated as the pIC50 – logD), distinguishes potency enhancements attributable
strictly to increased lipophilicity from those of targetꢀspecific interactions. Fit Quality (FQ), calculated as
the ligand efficiency (LE) of a compound adjusted for its number of heavy atoms, corrects the potency
values for the corresponding molecular weight.24 The utilization of FQ is particularly important when
comparing series and compounds of significant differences in size, as LE alone lends bias to smaller
compounds in such a comparison. These two metrics are especially useful for optimizing brain
availability, as increases in lipophilicity and MW diminish brain penetration via active efflux.25 The LipE
and FQ for compounds prepared via these approaches are plotted in Figure 5. This analysis revealed that
cyclic, secondary amines such as piperidines (blue dots) and azetidines (green diamonds) were preferred
cores for MAGL, whereas alternative ring systems or substitution patterns, especially those that were not
linear, were generally less efficient. Whereas azetidines and piperidines exhibited similar LipE values,
ACS Paragon Plus Environment

Page 7 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
with a number of exemplars of each registering between 5 and 6, a number of azetidines significantly
differentiated when gauging FQ.
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
Figure 5: Efficiency plots for MAGL inhibitors (a) Lipophilic efficiency plot comparing LogD (xꢀaxis)
vs MAGL pIC50 (b) Fit Quality plot comparing number of heavy atoms (xꢀaxis) vs Ligand Efficiency
(LE)
Highlighted compounds were screened against MAGL, as well as against FAAH as an initial,
albeit imperfect, surrogate for serine hydrolase selectivity to efficiently guide design. Specifically, 3ꢀ
substituted azetidines emerged as some of the most efficient linkers, as exemplified by 3ꢀ5 (Table 1).
Oxadiazole 3 displayed potent inhibition of MAGL (IC50 = 0.38 nM) nearly equivalent to that of the
literature tool KML29 (MAGL IC50 = 0.19 nM), albeit with significantly decreased molecular weight, as
reflected in its high FQ (0.97) and LipE. Pyrimidine 4 was slightly less potent, however its decreased
lipophilicity translated to an improvement in LipE (5.7) relative to 3 (LipE = 5.2). Moreover, while 3
maintained the previously observed selectivity for MAGL over FAAH (~1000ꢀfold), 4 exhibited
enhanced selectivity to >15,000ꢀfold. Piperazineꢀsubstituted 5 displayed similar potency (MAGL IC50 =
3.2 nM), efficiency (LipE = 5.5, FQ = 0.87), and selectivity over FAAH (781ꢀfold). Pyrazole 6, a slightly
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 48
1
2
3
4
less polar variant of oxadiazole 3, displayed modestly improved potency, and a subtle improvement in
5
6
LipE and FQ.
7
8
9
Table 1. In Vitro Data for 3ꢀ6.
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
Cpd
R =
MAGL IC50 (nM)a MAGL kinact/Ki
(Mꢀ1Sꢀ1)b
FAAH IC50 (nM)c LogD LipE FQ
0.38
840,174
420
4.2
3.0
5.2
5.7
0.97
0.89
3
4
1.9
121,478
>30,000
3.2
18,000
2,500
2.9
5.5
0.87
5
0.18
1,136,000
260
4.3
5.5
1.0
6
aIC50 values obtained from MAGL inhibition assay. kinact/KI ratio IC50 values obtained from FAAH
inhibition assay.
b
c
Collectively, these lead analogs 3ꢀ6 exemplified useful, efficient scaffolds for further
optimization. In general, the diversity in ring structures presented clear synthetic opportunities to grow
these subunits in a wide range of vectors. To guide design of the optimal group for the distal end of the
MAGL binding pocket, a crystal structure of 3 bound to the active site of MAGL was obtained (Figure 6).
The HFIP leaving group has clearly been displaced by the active site serine (Ser), resulting in a proteinꢀ
inhibitor carbamate adduct. The azetidine resides in the opening of the arachidonate port, illustrating the
narrow tunnel required and empirically suggested from our initial SAR efforts. The limited rotatable
bonds of 3 efficiently places the phenyl oxadiazolyl substituent into one of the distal pockets, in a similar
ACS Paragon Plus Environment

Page 9 of 48
Journal of Medicinal Chemistry
1
2
fashion to the proposed binding mode of compound 1.13 The hydrophobicity of the pocket surrounding
3
4
5
6
the oxadiazole and phenyl rings reinforces the need for minimal polarity in the tail region of the molecule.
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
Figure 6: Xꢀray crystal structure of 3 bound to the active site serine in MAGL (PDB ID 6AX1). The
residues making up the catalytic triad (Ser122, His269 and Asp239) have been shown in stick
representation. Hydrogen bonds are indicated by dashed lines. A glycerol molecule (grey), was modeled
into electron density seen in the exit tunnel. Glycerol was part of the buffer used to purify the protein and
was present in the crystallization drop.
Recognition of the pyrazole as an efficient linker off of the azetidine suggested modulation of the
pyrazole substituent to expand the SAR and understand if further improvements in efficiency were
available (Table 2). To gauge the efficiency of the pyrazole itself, the phenyl ring was stripped back and
replaced with a simple methyl (7), resulting in a significant attenuation of potency and decreased
efficiency. This result further reinforced the structural implication that a lipophilic tail was required.
Similarly, the substitution of an additional phenyl ring at the 3ꢀposition of the pyrazole (8) maintained
potency relative to 6, but the additional lipophilicity outpaced the corresponding potency gain, resulting in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 48
1
2
3
4
5
6
7
8
9
decreased efficiency. Difluorinated analog 9 did not impact MAGL potency or efficiency, but enhanced
selectivity over FAAH. Introduction of a pyrazine in place of the phenyl ring (10) resulted in a minor
decrease in MAGL potency, but the significant drop in lipophilicity resulted in a marked improvement in
LipE. Homologation to a benzylꢀsubstituted pyrazole loses efficiency at MAGL as well as selectivity
over FAAH. An attempt to decrease lipophilicity with the introduction of a 4ꢀsubstituted tetrahydropyran
(12) improved FAAH selectivity, but resulted in a 10X loss in MAGL potency. The anticipated decrease
in lipophilicity, however, offset the potency loss and resulted in similar LipE as the parent.
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 2. In Vitro Data for 7ꢀ12.
Cpd
R =
MAGL IC50
(nM)a
MAGL
FAAH IC50
(µM)a
LogD LipE FQ
kinact/Ki (Mꢀ1Sꢀ
1 b
)
39
4,047
>20
2.4
5.7
5.0
3.8
0.80
0.94
7
8
0.30
293,000
0.49
0.18
978,000
0.67
4.6
5.1
0.99
9
0.74
0.41
329,000
791,000
1.1
2.8
4.4
6.4
5.0
0.94
0.96
10
11
0.16
1.9
102,000
9.4
3.2
5.5
0.90
12
b
c
aIC50 values obtained from MAGL inhibition assay. kinact/Ki IC50 values obtained from FAAH
inhibition assay.
ACS Paragon Plus Environment

Page 11 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
With an exemplar such as 6 that was wellꢀoptimized for in vitro MAGL potency, we next
wanted to get a preliminary gauge of the broader selectivity against the other serine hydrolases. To this
end, 6 was screened against a panel of 42 serine hydrolases at 1 and 10 ꢀM concentrations of inhibitor
(see Table S2, Supporting Information). In this panel, 6 demonstrated significant inhibition (>70%) of
ABHD6, CES1, CES2, MAGL, and PLA2G7 at both concentrations tested and FAAH only at the top
concentration. To further explore the cellular potency and selectivity profile against other serine
hydrolases click chemistry activityꢀbased protein profiling (CCꢀABPP) was utilized.26 As such, an
alkyneꢀcontaining clickable probe (14, scheme 1) was prepared from bromophenyl pyrazole 13. After
installation of the alkyne via Sonagashira reaction with ethynyltrimethylsilane, subsequent deprotection of
both the alkyne and azetidine was followed by formation of the azetidinyl HFIP carbamate. Human brain
vascular pericytes were treated with increasing concentrations of 14, followed by cell lysis and click
chemistry with TAMRAꢀazide. The probeꢀlabelled proteins were visualized by inꢀgel fluorescence to
illustrate MAGL engagement and inhibition (Figure 7). Plotting the percentage of MAGL inhibition
demonstrated a cellular IC50 of 3 nM. Interesting, at concentrations below the IC50, the compound was
selective only for MAGL, however at higher concentrations, an offꢀtarget protein was labeled
corresponding to a molecular weight of ~40 kDa.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 48
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
Figure 7: Click chemistry activityꢀbased protein profiling in human brain vascular pericytes. Live cells
were treated at various concentrations of 14 for 1 h at 37 °C. (A) Inꢀgel fluorescence. (B) Normalized
MAGL intensity was fitted as a function of the concentration of 14.
With robust potency against recombinant MAGL as well as in human pericytes, compound 6 was
dosed in vivo in an effort to assess the relationship between drug levels and increases in 2ꢀAG levels in
the brain. As shown in Figure 8, a subcutaneous dose of 10 mg/kg to mouse elicited significant
elevations (~5ꢀfold, Figure 8a) in brain levels of 2ꢀAG out to an 8 h time point. Additionally, compound
6 showed free access to the CNS in mouse, as measured by an AUCꢀderived Cb,u/Cp,u ratio of 0.6. The
corresponding exposure for the 10 mg/kg dose exhibited a Tmax at 0.5 h (2 nM Cb,u) and had maintained
free levels above the measured IC50 to 8 h (Figure 8b). Further, a lower dose of 1 mg/kg provided a
dampened response, with a maximal 3ꢀfold increase of 2ꢀAG occurring at the 0.5 h time point, and no
clear elevation beyond that time point (Figure 8c). At this dose, the free brain exposure of compound 6
could only be measured at 0.5 h and was 0.02 nM, 100ꢀfold lower than the 10 mg/kg dose and well below
the in vitro IC50 (Figure 8d). Whereas the higher dose confirmed MAGL inhibition with treatment of
ACS Paragon Plus Environment

Page 13 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound 6, collectively, the data was somewhat confounding. First, the modest response at both doses
was surprising in comparison to previously run literature compounds with similar potency. Analysis of
the juxtaposed PK/PD alone would suggest that 6 was simply rapidly cleared, and the lack of exposure
precluded any opportunity for a 2ꢀAG response. Additionally, however, it was striking that the 1 mg/kg
dose showed 100ꢀfold lower free brain exposure than the 10 mg/kg dose. This nonꢀlinear PK suggested
that another phenomenon may be occurring, as well.
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
Mean Unbound Plasma and Brain Concentrations
of 6 @ 10 mg/kg SC Dose
10.00
Free Plasma (nM)
Free Cerebellum (nM)
1.00
0.10
0.01
0
5
10
15
Time (h)
(a)
(b)
Mean Unbound Plasma and Brain Concentrations
of 6 @ 1 mg/kg SC Dose
1.00
Free Plasma (nM)
Free Cerebellum
(nM)
0.10
0.01
0
1
2
3
Time (h)
(c)
(d)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 48
1
2
3
4
5
6
7
8
Figure 8: Effect of 6 on brain 2ꢀAG and corresponding exposure following subcutaneous administration
in wildꢀtype mice. (a) 2ꢀAG response following 10 mg/kg dose of 6 (b) Mean unbound plasma and brain
exposure (c) 2ꢀAG response following 1 mg/kg dose of 6 (d) Mean unbound plasma and brain exposure
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 an effort to understand whether this phenomenon was specific to the azetidine core itself or,
alternatively, the phenyl pyrazole, the corresponding piperidine 15 was prepared in similar fashion
(Figure 9a). Compound 15 retained similar in vitro activity against MAGL (IC50 = 2 nM) and serine
hydrolase selectivity panel (>70% inhibition of ABHD6, CES2, MAGL at both 1 and 10 ꢀM, and CES1,
FAAH, PLA2G7 only at 10 ꢀM, see Table S2, Supporting Information). Compound 15 was then
administered in vivo at the same doses as compound 6. Again, good brain penetration was observed, with
an AUCꢀderived Cb,u/Cp,u ratio of 0.5. At both doses a robust, persistent 2ꢀAG elevation was observed.
At the lower, 1 mg/kg, dose of 15 the 2ꢀAG response peaked at 2 h (Figure 9b), similar to the high dose of
6 albeit achieving a considerably lower Cmax (0.14 nM, Figure 9c). At the higher, 10 mg/kg, dose of 15
demonstrated a robust 2ꢀAG response out to 24 h (Figure 9d). Interestingly, at neither dose did central,
free levels of compound 15 exceed the IC50 (2 nM) measured in vitro, yet 2ꢀAG was significantly elevated
in both.
(a)
ACS Paragon Plus Environment

Page 15 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
Mean Unbound Plasma and Brain Concentrations
of 15 @ 1 mg/kg SC Dose
6
7
8
9
1.000
0.100
0.010
0.001
Free Plasma (nM)
Free Cerebellum (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
0
5
10
15
Time (h)
(b)
(c)
Mean Unbound Plasma and Brain
Concentrations of 15 @ 10 mg/kg SC Dose
1.00
0.10
0.01
Free Plasma (nM)
Free Cerebellum (nM)
0
10
20
30
Time (h)
(d)
(e)
Figure 9: (a) Structure of Compound 15. (bꢀc) Effect of 15 on brain 2ꢀAG following subcutaneous
administration in wildꢀtype mice of (a) 1 mg/kg and (b) 10 mg/kg doses.
In aggregate, the PK/PD data from 6 and 15 illustrated a disconnect in the exposure/2ꢀAG elevation
relationship for these compounds. Compound 6 achieved free exposures significantly higher than its in
vitro IC50, but showed only a marginal response, whereas 15 never exceeded its in vitro value but
exhibited a sustained 2ꢀAG elevation. It cannot be discounted that the lack of exquisite selectivity for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 48
1
2
3
4
5
6
7
8
9
MAGL as assessed by the serine hydrolase panel could potentially convolute the interpretation of the
data. However, the similar selectivity profile observed for both 6 and 15 minimizes offꢀtarget impacts as
cause for the discrepancy, as the same off target hydrolases are engaged to a similar extent. A number of
scenarios could potentially help to explain this PK/PD disconnect, including a) that the compounds were
not covalently inhibiting MAGL, and instead reverting to a paradigm wherein in vivo inhibition and
resulting pharmacodynamics response was driven by AUC coverage of the target, or b) that the inhibitors
were indeed covalently modifying MAGL, but that the resultant adducts varied in their hydrolytic
stability The combination of the cellular ABPP of 14 and the crystal structure of 3 covalently bound to
MAGL reinforced to us that the former scenario was unlikely. That same data, however, reinforces the
second scenario in that the azetidine carbamates were indeed forming a carbamate adduct with MAGL,
but that those complexes were less hydrolytically stable than those derived from the corresponding
piperidine. To this end, acylated azetidines have been shown to react differently than acyclic or larger
ring size amides due to decreased planarity in the amide moiety itself, conveyed by the ring strain
associated with the azetidine itself.27 Moreover, a comparison of the PK data suggested that in brain, the
azetidine carbamate had a much faster rate of clearance than the corresponding piperidine, which could
potentially be explained by initial reaction with MAGL, followed by a subsequent hydrolysis step. This,
in effect, would suggest that the target for inhibition was also its primary clearance pathway. Efforts to
further test this hypothesis, further deconvolute the observed PK/PD disconnect, and improve serine
hydrolase selectivity are underway and will be reported in due course.
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
CONCLUSION
In summary, highly efficient MAGL inhibitors were identified through a parallel medicinal chemistry
approach that highlighted the improved efficiency of piperidine and azetidineꢀderived carbamates.
Through monitoring improvements in FQ and LipE, efforts were focused on the optimization of 3ꢀ
substituted azetidines, including fiveꢀ and sixꢀmembered heterocycles. The incorporation of a pyrazole
served as an efficient linker to facilitate further growth into the distal lipophilic pocket. A crystal
ACS Paragon Plus Environment

Page 17 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
structure of 3 aided in subsequent design of inhibitors and suggested minimizing the exposed polarity on
the inhibitor. Replacement of the oxadiazole with a pyrazole, substituted with either a phenyl (6) or
pyrazine (10) ring maximized potency and LipE, respectively. The cellular potency an exemplar of these
compounds was confirmed using ABPP. An in vivo experiment monitoring the effect of a key exemplar
on 2ꢀAG elevation confirmed inꢀvivo MAGL inhibition, but a modest response suggested that the
azetidineꢀderived analogs were exhibiting dampened PD effects than would be expected for a covalent
modification of MAGL, whereas the corresponding piperidine 15 demonstrated a robust 2ꢀAG elevation.
Further efforts to understand this phenomenon are ongoing and will be reported in due course.
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
Biology
In Vitro Pharmacology Assessment of MAGL inhibition (MAGL IC50): Assessment of MAGL inhibition
utilizes human recombinant Monoacylglycerol Lipase and the fluorogenic substrate 7ꢀhydroxycoumarinyl
arachidonate (7ꢀHCA, Biomol STꢀ502). 400 nL of a test compound at decreasing concentration (ranging
from 150 µM down to 1.5 nM) was spotted into a 384ꢀwell back plate (PerkinElmer, 6007279) using a
Labcyte Echo, followed by addition of 10 µL of MAGL enzyme in assay buffer (50mM HEPES, pH 7.4,
100 mM NaCl, 5 mM MgCl2, 0.1% Triton Xꢀ100 and 25% glycerin). An equal volume of 7ꢀHCA in
assay buffer with 10% DMSO was added either immediately (T = 0 min) or after a 30 min incubation (T
= 30 min) to initiate the reaction. The final concentration of MAGL enzyme was 88 pM and 7ꢀHCA
substrate was 5 µM. After these dilutions, the final concentration of the test compound ranged from 3 µM
to 0.03 nM. The reaction was allowed to progress for 60 minutes, after which the plate was read at an
Ex/Em of 340/465. Percent inhibitions were calculated based on control wells containing no compound
(0% inhibition) and a control compound (e.g., a MAGL inhibitor whose activity is known or was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 48
1
2
3
4
5
6
previously reported in the literature, such as one with about 100% inhibition). IC50 values were
generated based on a four parameter fit model using ABASE software from IDBS.28
7
8
9
Assessment of rate of MAGL inactivation (MAGL kinact/KI): To measure MAGL inactivation, the same
protocol for the (T = 0 min) MAGL inhibition IC50 assay was performed with data collected every
minute to acquire enzyme progress curves at decreasing concentrations of compound. Kobs values were
calculated from this data and kinact/KI ratios were determined from a plot of Kobs values vs. compound
concentrations.
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
Assessment of FAAH inhibition: Assessment of FAAH inhibition utilizes human recombinant Fatty Acid
Amide Hydrolase Lipase and the fluorogenic substrate, ArachidonoylꢀAMC (Sigma A6855). 400 nL of
a test compound at decreasing concentration was spotted into a 384ꢀwell back plate (PerkinElmer,
6007279) using a Labcyte Echo, followed by addition of 10 ꢁl of FAAH enzyme (Cayman 10010183) in
assay buffer (50 mM Tris, pH 9.0, 1 mM EDTA). After a 30 minute incubation at room temperature, 10
ꢁL of ArachidonylꢀAMC was added in assay buffer with 16% DMSO. Final concentration of FAAH
enzyme was 0.0125 Units and the AAMC substrate was used at a concentration of 5 ꢁM. The final
concentration of the test compound ranged from 30 µM to 0.3 nM. The reaction was allowed to progress
for 60 minutes, after which the plate was read at an Ex/Em of 355/460. Percent inhibitions were
calculated based on control wells containing no compound (0% inhibition) and a control compound at a
concentration known to inhibit FAAH to 100%. IC50 values were generated based on a four parameter fit
model using ABASE software from IDBS.
Cellular Pharmacology Human brain vascular pericytes and pericyte growth supplement were
purchased from ScienCell Research Laboratories. Media, other supplements for cell culture, and reagents
for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDSꢀPAGE) were purchased from
Thermo Fisher Scientific unless otherwise noted. TAMRA azide was purchased from Lumiprobe. DMSO
solutions of compound 14 were prepared and aliquots were stored at ꢀ20 °C.
ACS Paragon Plus Environment

Page 19 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Cell Culture and Treatment: Human brain vascular pericytes were cultured in Dulbecco's Modified Eagle
Medium/Nutrient Mixture Fꢀ12 (DMEM/F12), supplemented with 5% heatꢀinactivated fetal bovine serum
(FBS), 1% pericyte growth supplement, and 100 U/mL penicillinꢀstreptomycin in a humidified
environment maintained with 5% CO2 at 37 °C. For labeling with probe 14, cells were cultured on 6ꢀwell
plates overnight, and 1000x stock solutions of probe 14 in DMSO was added to the growth media, and
cells were maintained for 1 h at 37 °C. At the end of treatment, cells were washed with 1 mL of cold
phosphateꢀbuffered saline (PBS), and was scrapped in 1 mL of cold PBS. Cells were pelleted by
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
ꢀ
centrifuging at 10,000×g at 37 °C for 1 min. The supernatant was removed and cell pellets were kept at
80°C until further analysis.
Click Chemistry Activity-Based Protein Profiling: To the cell pellets were added PBS containing 0.25%
SDS, and the cells were lysed with gentle sonication. The proteome concentration was measured with
bicinchoninic acid (BCA) assay. The proteomes were normalized to 1.0 mg/mL and subject to Cuꢀ
catalyzed click chemistry with TAMRA azide following a reported protocol.29 The reaction mixtures were
diluted with lithium dodecyl sulfate (LDS) sample buffer and NuPAGE reducing agent, and were
analyzed with 1.0 mm NuPAGE 4ꢀ12% bisꢀtris 15ꢀwell gels in 2ꢀ[Nꢀmorpholino]ethanesulfonic acid
(MES) running buffer. TAMRA fluorescence signals were collected on a Typhoon FLA 9500
Biomolecular Imager (GE Healthcare) with 532 nm laser excitation and ≥575 nm longꢀpass emission
filter.
Data Analysis: Inꢀgel fluorescence data were visualized with ImageJ software (v1.47, NIH). Fluorescence
intensity was quantified with background subtraction using ImageStudio Lite (v4.0.21, LIꢀCOR). The
intensity of the two MAGL bands were averaged, and data from two biological replicates were plotted as
mean ± standard deviation in GraphPad Prism (v7.02) and were fit to an inhibitor doseꢀresponse curve to
obtain the cellular IC50.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 48
1
2
3
4
5
6
7
8
In Vivo Experiments All procedures performed on animals in this study were in accordance with
established guidelines and regulations, and were reviewed and approved by the Pfizer (or other )
Institutional Animal Care and Use Committee.
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
Assessment of 2-AG Elevation: C57Bl6 mice (7 weeks old) were housed 4 per cage in an environmentally
controlled animal facility (12 hour light/dark cycle from 6:00 AM to 6:00 PM; temperature maintained
between 20.5 and 22.0 °C; relative humidity maintained between 50ꢀ65%). Food and water were
available ad libitum throughout study. Mice were subcutaneously administered compound 6, 14, or
vehicle (DMSO:Cremophor:Saline at a volume ratio of 5:5:90). Plasma and brain samples of 6 and 14
treated mice were collected at 0.5, 1, 2, 4, 8, 12, and 24 hours post dose (3 mice per time point). Plasma
and brain samples of vehicle treated mice were collected at 1 hour post dose. Blood was collected by first
anesthetizing mice with isoflurane followed by cardiac puncture. Whole blood was placed into tubes
containing NaF and K2EDTA (0.5 M, 20 ꢀL per 1 mL blood) to inhibit plasma hydrolase activity and as
an antiꢀcoagulant, respectively. Blood was centrifuged at 4,000 rcf for 5 min to harvest plasma.
Immediately following blood sampling mice were euthanized by cervical dislocation and their brains
were removed. Brain and cerebellum were immediately frozen in liquid nitrogen and stored at ꢀ80°C until
analysis.
Pharmacodynamic measurement. Plasma samples were prepared for 2 AG measurement on wet ice.
Briefly, samples were spiked with IS2 (internal standard 2) solution [1 ꢀg/mL d5 2 AG (deuterated 2
arachidonoylglycerol) (Cayman Chemical Ann Arbor, Michigan) in ACN] followed by precipitation and
organic extraction in EtOAC/Hexane (v:v 50:50). Samples were then centrifuged at 15,700 rcf for 15
min. Supernatant was removed and dried under nitrogen gas at room temperature. Samples were then
reconstituted in ACN and directly injected for LC MS/MS analysis. Calibration standards and quality
control samples were prepared using the same methodology. The 2 AG calibration curve consisted of 2
500 ng/mL of 2 AG in water/PMSF/buffer [100 mM NH4OAc in water (pH 2.0 adjusted with formic
acid)] (v:v:v 100:5:20). For 2 AG measurement samples were directly injected onto a C 18 UPLC column
ACS Paragon Plus Environment

Page 21 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
held at 65 °C at a flow rate of 0.5 mL/min. Retention time for 2 AG was 2.71 min. Retention time for
d5ꢀ2ꢀAG was 2.70 min. Mass spectrometry was run in positive electrospray ionization (ESI) mode with
selected reaction monitoring (SRM) detection. The 2 AG m/z was 379.4/287.3. The d5ꢀ2ꢀAG m/z was
384.4/287.4.
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
Brain samples were prepared on wet ice. Briefly, brain homogenates were prepared by homogenizing
brain with 3 volumes (w:v) of homogenization buffer (1% PMSF and 5% 100 mM NH4OAc in water pH
2.0, adjusted with formic acid). The brain homogenates were then spiked with IS2 [1 ꢀg/mL d5 2 AG,
(deuterated 2 arachidonoylglycerol) (Cayman Chemical Ann Arbor, Michigan) in ACN] solution
followed by precipitation and organic extraction in ACN. Samples were then centrifuged at 15,700 rcf for
15 min at 4°C. Supernatant was removed from each sample directly injected for LC MS/MS analysis. A
d5 2 AG internal standard (Cayman Chemical Ann Arbor, Michigan) was added to each sample.
Calibration standards and quality control samples were prepared using the same methodology. The 2 AG
calibration curve consisted of 50 50,000 ng/mL 2 AG in homogenization buffer. For 2 AG measurement
samples were directly injected onto a C 18 UPLC column held at 50°C at a flow rate of 0.5 mL/min.
Retention time for 2 AG was 3.30 min. Retention time for d5 2 AG was 3.29 min. Mass spectrometry
was run in positive ESI mode with SRM detection. The 2 AG m/z was 379.4/287.3
Analysis of compound in plasma and brain and pharmacokinetic parameters All samples were
quantified via HPLCꢀMS/MS. Cerebellum samples were thawed, diluted 1:4 (w/v) with water,
and homogenized using an Hꢀspeed dispersator. Standards and controls were prepared in a
similar manner using a brain homogenate prepared from untreated animals. Plasma and brain
homogenate samples were prepared for analysis by deproteination with methanol/acetonitrile
(1:1, v/v) containing an internal standard. The sample volume for analysis was 10 ꢀL and 50 ꢀL
for the plasma and brain matrices, respectively. The samples were precipitated with 200 ꢀl of
methanol/acetonitrile, vortexed, and centrifuged. Following centrifugation, the supernatant was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 48
1
2
3
4
5
6
7
8
9
transferred to a clean 96ꢀwell injection block and diluted 40x with water containing 0.1% formic
acid. A sample volume of 10 ꢀL for the plasma samples and 5 ꢀL for the brain samples was
subsequently injected on the LCꢀMS/MS for analysis. The lower limit of quantification (LLOQ)
was 1.0 ng/ml for plasma and 0.5 ng/mL for brain homogenate (conversion to 2.5 ng/g) samples.
The upper limit of quantification (ULOQ) was 1000 ng/mL for plasma and 500 ng/mL for brain
homogenate (conversion to 2500 ng/g). Pharmacokinetic parameters were determined by nonꢀ
compartmental analysis using Watson LIMS version 7.5.
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
X-ray structure of MAGL bound to 3: Synthetic DNA encoding fullꢀlength human MAGL (residues 1ꢀ
303), with an Nꢀterminal 6XꢀHis Tag and a Cꢀterminal Strepꢀtag was cloned into E. coli (Rosetta strain)
and expressed and purified essentially as described previously 30. The protein was purified by a
combination of nickel ion and streptactin affinity chromatography and concentrated to 50 mg/mL in a
final buffer consisting of 15 mM HEPES pH 8.2, 2 mM TCEP, and 10% glycerol, with hexaethylene
glycol monododecyl ether added to 0.1 mM. Crystallization was carried out in sitting drops with 0.30 ꢀL
of protein at 20 mg/mL added to 0.30 ꢀL of reservoir solution (0.07 M sodium cacodylate pH 5.1ꢀ5.9 and
33ꢀ51% MPD).
Apo crystals obtained from these drops were harvested and transferred to a cryoprotectant soaking
solution that consisted of 70 mM sodium cacodylate pH 5.1; 10% MPD; 30% PEGꢀMMEꢀ2K with the
addition of 1 mM Compound 3. Crystals were left to soak overnight (12ꢀ18hrs) and flashꢀfrozen in liquid
nitrogen for data collection.
Xꢀray diffraction data to 2.26 Å resolution were collected at 100K with radiation of wavelength 1 Å at
beamline 17ꢀID of the Advanced Photon Source at Argonne, on a Pilatusꢀ6M detector. Data were
processed using autoPROC31 with XDS32 for data integration and the program Scala from the CCP4
Suite33 for merging and scaling. The structure was solved by molecular replacement using PDB entry
ACS Paragon Plus Environment

Page 23 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
3HJU30 as the search model, and refined with autoBUSTER.34 Model building and fitting was done using
Coot.35 Data collection and refinement statistics are given in Table S1.
6
7
8
9
Chemistry
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
General Methods. Solvents and reagents were of reagent grade and were used as supplied by the
manufacturer. All reactions were run under a N2 atmosphere. Organic extracts were routinely
dried over anhydrous sodium sulfate. Concentration refers to rotary evaporation under reduced
pressure. Chromatography refers to flash chromatography using disposable RediSepRf 4 to 120
g silica columns or Biotage disposable columns on a CombiFlash Companion or Biotage
Horizon automatic purification system. Microwave reactions were carried out in a SmithCreator
microwave reactor from Personal Chemistry. Purification by massꢀtriggered HPLC was carried
out using Waters XTerra PrepMS C18 columns, 5 ꢀm, 30 x 100 mm. Compounds were preꢀ
salted as TFA salts and diluted with 1 mL dimethylsulfoxide. Samples were purified by mass
triggered collection using a mobile phase of 0.1% trifluoroacetic acid in water and acetonitrile
with a starting gradient of 100% aqueous to 100% acetonitrile over 10 minutes at a flow rate of
20 mL/minute.
Elemental analyses were performed by QTI, Whitehouse, NJ. All target
compounds were analyzed using ultra high performance liquid chromatography /ultra violet/
evaporative light scattering detection coupled to time of flight mass spectrometry
(UHPLC/UV/ELSD/TOFMS). Unless otherwise noted, all tested compounds were found to be
>95% pure by this method.
UHPLC/MS Analysis. The UHPLC was performed on a Waters ACQUITY UHPLC system
(Waters, Milford, MA), which was equipped with a binary solvent delivery manager, column
manager, and sample manager coupled to ELSD and UV detectors (Waters, Milford, MA).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 48
1
2
3
4
5
6
7
8
9
Detection was performed on a Waters LCT premier XE mass spectrometry (Waters, Milford,
MA). The instrument was fitted with an Acquity BEH (Bridged Ethane Hybrid) C18 column (30
mm × 2.1 mm, 1.7 ꢁm particle size, Waters, Milford, MA) operated at 60 ° C.
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
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ(3ꢀphenylꢀ1,2,4ꢀoxadiazolꢀ5ꢀyl)azetidineꢀ1ꢀcarboxylate (3)
Step 1: To a solution of Nꢀhydroxybenzenecarboximidamide (6.5 g, 48 mmol), 1ꢀ(tertꢀ
butoxycarbonyl)azetidineꢀ3ꢀcarboxylic acid (11.5 g, 57 mmol) and HATU (21.8g, 57 mmol) in CH2Cl2
(80 mL) at room temperature was added DIPEA (18.5 g, 143 mmol), and the resulting mixture was
stirred for 16 h. The mixture was then portioned with water (40 mL x 3) and then saturated sodium
chloride (40 mL), dried over sodium sulfate, filtered and concentrated in vacuo. The residue was
dissolved in NMP (80 mL) and heated to 140 ° C for 5 h. The mixture was cooled to room temperature,
poured onto water and extracted with ethyl acetate (50 mL x 3). The combined organic layers were
washed with water and saturated sodium chloride, dried over sodium sulfate, and concentrated in vacuo.
The residue was purified on a silica gel column (gradient: petroleum ether to 1:1 petroleum ether:ethyl
acetate) to afford 10.6 g (74% yield) of tertꢀbutyl 3ꢀ(3ꢀphenylꢀ1,2,4ꢀoxadiazolꢀ5ꢀyl)azetidineꢀ1ꢀ
carboxylate as a yellow oil
.Step 2: To a mixture of tertꢀbutyl 3ꢀ(3ꢀphenylꢀ1,2,4ꢀoxadiazolꢀ5ꢀyl)azetidineꢀ1ꢀcarboxylate (3.5 g, 11.6
mmol) in ethyl acetate (10 mL), was added a solution of HCl in ethyl acetate (4 M, 40 mL), and the
reaction was stirred at room temperature for 30 min. The reaction was then concentrated in vacuo, to give
2.62 g (95%) of the HCl salt of 5ꢀ(azetidinꢀ3ꢀyl)ꢀ3ꢀphenylꢀ1,2,4ꢀoxadiazole as a white solid.
Step 3: To a solution of bis(trichloromethyl) carbonate (252 mg, 0.84 mmol ) in CH2Cl2(40 mL) was
added 1,1,1,3,3,3ꢀhexafluoropropanꢀ2ꢀol (426mg, 2.52 mmol) and DIPEA ( 330 mg, 2.52 mmol ) at 0℃
under N2. The reaction mixture was stirred at 20 ℃ for 7 h. A solution of 5ꢀ(azetidinꢀ3ꢀyl)ꢀ3ꢀphenylꢀ
1,2,4ꢀoxadiazole hydrochloride (600 mg, 2.52 mmol) and DIPEA ( 1.3 g, 10 mmol) in CH2Cl2 (10 mL)
was added to the mixture and then was stirred at 20 ℃ for 16 h. The reaction mixture was then diluted
ACS Paragon Plus Environment

Page 25 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with CH2Cl2 and washed with water (50 mL x 3) and saturated sodium chloride (50 mL), dried over
sodium sulfate and concentrated in vacuo to give the crude product. The residue was purified by silica
gel chromatography (petroleum ether:ethyl acetate, 5:1) to give 470 mg (47%) of 1,1,1,3,3,3ꢀ
hexafluoropropanꢀ2ꢀyl 3ꢀ(3ꢀphenylꢀ1,2,4ꢀoxadiazolꢀ5ꢀyl)azetidineꢀ1ꢀcarboxylate (3) as a clear oil. 1H
NMR (400 MHz, CD3OD) δ ppm 8.11ꢀ 8.07 (m, 2 H), 7.59 ꢀ 7.49 (m, 3 H), 6.12 (dt, J = 12.5 and 6.3 Hz,
1 H), 4.68 ꢀ 4.53 (m, 2 H), 4.51 ꢀ 4.38 (m, 2 H), 4.38 ꢀ 4.30 (m, 1 H),. 13C NMR (101MHz, CDCl3) δ
177.9, 168.7, 151.4, 131.5, 129.0, 127.5, 126.2, 120.5 (q, 1JCF =283.2 Hz), 67.8 (sep, 2JCF =34.5 Hz), 54.1,
53.4, 26.3. LC/MS (ESI, m/z) 396.0 [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
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ(2ꢀmorpholinopyrimidinꢀ4ꢀyl)azetidineꢀ1ꢀcarboxylate (4)
To a solution of triphosgene (38 mg, 0.13 mmol ) in CH2Cl2 (5 mL) was added 1,1,1,3,3,3ꢀ
hexafluoropropanꢀ2ꢀol (66 mg, 0.39 mmol) and DIPEA ( 50 mg, 0.39 mmol ) at 0 ℃ under N2. The
reaction mixture was stirred at 20℃ for 4 h. Added 4ꢀ(4ꢀ(azetidinꢀ3ꢀyl)pyrimidinꢀ2ꢀyl)morpholine (100
mg, 0.39 mmol) and DIEA ( 202 mg, 1.56 mmol) in CH2Cl2 (5 mL) and the reaction mixture was stirred
at 20℃ for 16 h. The reaction mixture was partitioned between CH2Cl2,(20 mL) and water (30 mL x 3),
and then washed with saturated sodium chloride (30 mL), dried over sodium sulfate, concentrated in
vacuo to give the crude product which was purified by prep TLC on silica gel (petroleum ether:ethyl
acetate , 2:1) to afford 57 mg (35%) of 1,1,1,3,3,3ꢀhexafluoropropanꢀ2ꢀyl 3ꢀ(2ꢀmorpholinopyrimidinꢀ4ꢀ
1
yl)azetidineꢀ1ꢀcarboxylate (4) as a white solid. H NMR (400 MHz, CD3OD) δ ppm 8.26 (d, J = 5.1 Hz,
1 H), 6.54 (d, J = 4.8 Hz, 1 H), 6.09 (spt, J = 6.4 Hz, 1 H), 4.47 ꢀ 4.34 (m, 2 H), 4.34 ꢀ 4.20 (m, 2 H), 3.87
ꢀ 3.82 (m, 1 H), 3.81ꢀ3.70 (m, 8 H). LC/MS (ESI, m/z) 415.3 [M + H]+
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ[4ꢀ(pyrimidinꢀ2ꢀyl)piperazinꢀ1ꢀyl]azetidineꢀ1ꢀcarboxylate (5)
To a solution of 2ꢀ[4ꢀ(azetidinꢀ3ꢀyl)piperazinꢀ1ꢀyl]pyrimidine36 (100 mg, 0.30 mmol) and DIPEA (196
mg, 1.5 mmol) in CH2Cl2 (6 mL) was added a solution of triphosgene, 1,1,1,3,3,3ꢀhexafluoropropanꢀ2ꢀol,
DIPEA, and DMAP in CH2Cl2 dropwise at 0 °C. The mixture was stirred at 16 °C for 3 h. The mixture
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 48
1
2
3
4
5
6
7
8
9
was concentrated in vacuo and purified by preparative high performance liquid chromatography to give
21 mg (17 %) of 1,1,1,3,3,3ꢀhexafluoropropanꢀ2ꢀyl 3ꢀ[4ꢀ(pyrimidinꢀ2ꢀyl)piperazinꢀ1ꢀyl]azetidineꢀ1ꢀ
carboxylate (5) as a white solid. 1H NMR (400 MHz, CDCl3) δ 8.31 (d, J=4.7 Hz, 2 H), 6.51 (t, J=4.7
Hz, 1 H), 5.65 (sep, J=6.2 Hz, 1 H), 4.23 ꢀ 3.97 (m, 4 H), 3.93 ꢀ 3.79 (m, 4 H), 3.29 ꢀ 3.16 (m, 1 H), 2.44
– 2.39 (m, 4 H). 13C NMR (101MHz, CDCl3) δ 161.5, 157.7, 151.7, 120.6 (q, 1JCF =283.2 Hz), 110.2, 67.6
(sep, 2JCF =34.5 Hz), 54.2, 54.0, 53.3, 49.4, 43.2. HRMS calculated for C15H17F6N5O2 [M+H]+ 414.1359,
found 414.1361.
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
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ(1ꢀphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (6)
Step 1: tertꢀbutyl 3ꢀ(1ꢀphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate A mixture of tertꢀbutyl 3ꢀ
[(2E)ꢀ3ꢀ(dimethylamino)propꢀ2ꢀenoyl]azetidineꢀ1ꢀcarboxylate (800 mg, 3.15 mmol) and phenyl
hydrazine (340 mg, 3.15 mmol) in EtOH (10 mL) was heated to reflux and stirred for 16 h. The reaction
mixture was concentrated in vacuo and purified by silica gel chromatography to give a mixture of tertꢀ
butyl 3ꢀ(1ꢀphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (200 mg, 21.25%) and tertꢀbutyl 3ꢀ(1ꢀphenylꢀ
1Hꢀpyrazolꢀ5ꢀyl)azetidineꢀ1ꢀcarboxylate (280 mg, 29.8%) as a brown solid. Data for tertꢀbutyl 3ꢀ(1ꢀ
1
phenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate: H NMR (400 MHz, CDCl3) δ ppm 7.82 (d, J = 2.4 Hz,
1 H), 7.65 ꢀ 7.57 (m, 2 H), 7.37 (t, J = 7.9 Hz, 2 H), 7.24 ꢀ 7.18 (m, 1 H), 6.37 (d, J = 2.4 Hz, 1 H), 4.35 ꢀ
4.19 (m, 2 H), 4.04 (dd, J = 8.4 and 6.2 Hz, 2 H), 3.83 (s, 1 H), 1.45 ꢀ 1.32 (m, 9 H).
Step 2: 3ꢀ(Azetidinꢀ3ꢀyl)ꢀ1ꢀphenylꢀ1Hꢀpyrazole To a solution of tert-butyl 3ꢀ(1ꢀphenylꢀ1Hꢀpyrazolꢀ3ꢀ
yl)azetidineꢀ1ꢀcarboxylate (200 mg, 0.67 mmol) in EtOAc (0.5 mL) was added a solution of HCl in
MeOH (1 mL, 4 mmol, 4 M) dropwise at 0 °C. The resulting mixture was stirred at room temperature (25
°C) for 1 h. The reaction was concentrated in vacuo and the residue was triturated with MTBE (20 mL x
3) and evaporated to give the 150 mg (97%) of the crude 3ꢀ(azetidinꢀ3ꢀyl)ꢀ1ꢀphenylꢀ1Hꢀpyrazole HCl salt
as a dark yellow solid, which was used in the next step directly.
Step 3: 1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ(1ꢀphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (6)
ACS Paragon Plus Environment

Page 27 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A solution of hexafluoroisopropanol (108 mg, 0.64 mmol), DMAP (8.3 mg, 0.06 mmol) and DIPEA (99
mg, 0.77 mmol) in CH2Cl2 (5 mL) was added dropwise to a solution, preꢀcooled (0 °C, under an N2
atmosphere), of triphosgene (63 mg, 0.21 mmol) in CH2Cl2 (3 mL). The resulting mixture was stirred at
25 °C for 16 h. This solution was then added to a stirring solution of 3ꢀ(azetidinꢀ3ꢀyl)ꢀ1ꢀphenylꢀ1Hꢀ
pyrazole hydrochloride (150 mg, 0.64 mmol) and DIPEA (248 mg, 1.92 mmol) in CH2Cl2 (2 mL). The
resulting mixture was stirred at 25 °C for 6 h.. The reaction mixture was quenched with water (20 mL),
and extracted with EtOAc (20 mL x 3). The combined organic layer was dried over sodium sulfate,
filtered, concentrated in vacuo, and purified by prepꢀHPLC to give 50 mg (20%) of 1,1,1,3,3,3ꢀ
hexafluoropropanꢀ2ꢀyl 3ꢀ(1ꢀphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (6) as a white solid. 1H
NMR (400 MHz, CDCl3) δ ppm 7.91 (d, J = 2.5 Hz, 1 H), 7.67 (d, J = 8.0 Hz, 2 H), 7.46 (s, 2 H), 7.26 (s,
1 H), 6.42 (d, J = 2.5 Hz, 1 H), 5.70 (dt, J = 12.5 and 6.2 Hz, 1 H), 4.59ꢀ4.43 (m, 2 H), 4.42 ꢀ 4.25 (m, 2
H), 4.13 – 3.99 (m, 1 H). 13C NMR (101MHz, CDCl3 ) δ 153.6, 151.8, 139.9, 129.5, 128.1, 126.6, 120.6
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
2
(q, JCF =282.4 Hz), 119.1, 105.3, 67.6 (sep, JCF =34.5 Hz), 56.2, 55.5, 28.2. HRMS calculated for
C16H13F6N3O2 [M+H]+ 396.0777, found 396.0766.
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (7) To a
solution of triphosgene (89 mg, 0.3 mmol) in CH2Cl2 (1 mL) was added a solution of 1,1,1,3,3,3ꢀ
hexafluoropropanꢀ2ꢀol (153 mg, 0.9 mmol), DIPEA ( 118 mg, 0.9 mmol) and DMAP ( 11 mg, 0.09
mmol) in CH2Cl2 (1 mL) dropwise at ꢀ10 ℃. To this solution was added a solution of 3ꢀ(azetidinꢀ3ꢀyl)ꢀ1ꢀ
methylꢀ1Hꢀpyrazole (125 mg, 0.9 mmol) and DIPEA (353 mg, 2.7 mmol) in CH2Cl2 (1 mL) and the
resulting mixture was stirred at 15 ℃ overnight. The reaction mixture was then filtered, concentrated and
purified by prep TLC (30 % ethyl acetate in petroleum ether) to give 20 mg (7%) of 1,1,1,3,3,3ꢀ
hexafluoropropanꢀ2ꢀyl 3ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (7). 1H NMR (400 MHz,
CD3OD) δ ppm 7.82 ꢀ 7.29 (m, 1 H), 6.24 (d, J = 2.2 Hz, 1 H), 6.05 (dquin, J = 12.7 and 6.4 Hz, 1 H),
4.53 ꢀ 4.35 (m, 2 H), 4.24 ꢀ 4.08 (m, 2 H), 3.94 (tt, J = 8.8 and 6.3 Hz, 1 H), 3.85 (s, 3 H) 13C NMR
2
(101MHz, CD3OD) δ 153.75, 153.34, 133.44, 104.63, 68.88(sep, JCF = 34.06 Hz), 57.69, 57.01, 38.87,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 48
1
2
3
4
5
29.25, CF3 carbons not observed. HRMS calculated for C11H11F6N3O2 [M+H]+ 332.0828, found
332.0825.
6
7
8
9
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ(1,5ꢀdiphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (8): To
a solution of triphosgene (31 mg, 0.11 mmol) in CH2Cl2 (10 mL) was added 1,1,1,3,3,3ꢀ
hexafluoroisopropanol (54 mg, 0.32 mmol) and DIPEA (41 mg, 0.32 mmol ) at 0 °C under N2. The
reaction mixture was stirred at 27 °C for 7 h. To this solution was added a solution of 3ꢀ(azetidinꢀ3ꢀyl)ꢀ
1,5ꢀdiphenylꢀ1Hꢀpyrazole (100 mg, 0.32 mmol) and DIPEA (166 mg, 1.3 mmol) in CH2Cl2 (5 mL) and
then the reaction mixture was stirred at 27 °C for 16 hours. The reaction mixture was then partitioned
between CH2Cl2 (15 mL) and water (50 mL), and the organic layer was washed with water (50 mL x 3)
and saturated sodium chloride (50 mL), dried over sodium sulfate, filtered, and concentrated in vacuo to
give the crude product, which was purified by prep TLC (petroleum ether:EtOAc 6:1) to give 71 mg
(46%) of 1,1,1,3,3,3ꢀhexafluoropropanꢀ2ꢀyl 3ꢀ(1,5ꢀdiphenylꢀ1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (8)
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
as a white solid. H NMR (400 MHz, CD3OD) δ ppm 7.64 ꢀ 7.01 (m, 11 H), 6.71ꢀ6.53 (m, 1 H), 6.19ꢀ
5.97 (m, 1 H), 4.61ꢀ4.44 (m, 2 H), 4.38ꢀ4.22 (m, 2 H), 4.14 ꢀ3.98 (m, 1 H). 13C NMR (101MHz, CDCl3)
1
δ 152.6, 151.7, 144.5, 139.8, 130.2, 129.0, 128.6, 128.5, 128.4, 127.6, 125.2, 120.6 (q, JCF =282.4 Hz),
105.3, 67.6 (sep, 2JCF =34.5 Hz), 56.2, 55.6, 28.2. LC/MS (ESI, m/z) 470.0 [M + H]+
1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ[1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazolꢀ3ꢀyl]azetidineꢀ1ꢀ
carboxylate (9)
Step 1: tertꢀButyl 3ꢀ[1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazolꢀ3ꢀyl]azetidineꢀ1ꢀcarboxylate (A) A mixture
of tertꢀbutyl 3ꢀ(1Hꢀpyrazolꢀ3ꢀyl)azetidineꢀ1ꢀcarboxylate (200 mg, 0.9 mmol), 3,4ꢀdifluorophenyl boronic
acid (170 mg, 1.1 mmol), Cu(OAc)2 (244 mg, 1.3 mmol) and pyridine (213 mg, 2.7 mmol) in CH2Cl2 (30
mL) was stirred at 40 °C for 20 h. The reaction mixture was then filtered and the filtrate was diluted with
water (20 mL) and extracted with CH2Cl2 (20 mL x 3). The combined organic layer was washed with
water (20 mL), saturated sodium chloride (20 mL), dried over sodium sulfate, concentrated in vacuo and
ACS Paragon Plus Environment

Page 29 of 48
Journal of Medicinal Chemistry
1
2
3
4
purified by silica gel chromatography (ethyl acetate/petroleum ether, gradient 0 to 10%,) to give 150 mg
5
6
(50%) of tertꢀbutyl 3ꢀ[1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazolꢀ3ꢀyl]azetidineꢀ1ꢀcarboxylate (A) as a white
7
8
1
solid. H NMR (400 MHz, CDCl3) δ ppm 7.82 (d, J = 2.5 Hz, 1 H), 7.56 (ddd, J = 11.2, 6.9 and 2.5 Hz, 1
9
H), 7.40ꢀ 7.33 (m, 1 H), 7.28 ꢀ 7.11 (m, 1 H), 6.44 (d, J = 2.5 Hz, 1 H), 4.36 ꢀ 4.26 (m, 2 H), 4.13 ꢀ 4.05
(m, 2 H), 3.87 (tt, J = 8.7, 6.1 Hz, 1 H), 1.50 ꢀ 1.41 (m, 9 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
Step 2: 3ꢀ(Azetidinꢀ3ꢀyl)ꢀ1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazole (B) To a solution of A (150 mg, 0.45
mmol) in CH2Cl2 (0.5 mL) was added a solution of HCl in EtOAc (0.56 mL 2.2 mmol, 4 M) dropwise at
0 °C. The resulting mixture was stirred at room temperature for 1 h, and then concentrated in vacuo. The
residue was triturated with TBME (10 mL x 3) and then concentrated to give 120 mg (99%) of crude 3ꢀ
(azetidinꢀ3ꢀyl)ꢀ1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazole (B) as a white solid, which was used in the next step
without purification.
Step 3: 1,1,1,3,3,3ꢀHexafluoropropanꢀ2ꢀyl 3ꢀ[1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazolꢀ3ꢀyl]azetidineꢀ1ꢀ
carboxylate (9) A solution of hexafluoroisopropanol (66.0 mg, 0.39 mmol), DMAP (4.8 mg, 0.04 mmol)
and DIPEA (61 mg, 0.47 mmol) in CH2Cl2 (5 mL) was added dropwise to a solution of triphosgene (38.5
mg, 0.13 mmol) in CH2Cl2 (3 mL) at 0 °C under a nitrogen atmosphere. The resulting mixture was stirred
at 30 °C for 16 h. The solution was then added to a solution of B (120 mg, 0.39 mmol) and DIPEA (152
mg, 1.2 mmol) in CH2Cl2 (2 mL). The resulting mixture was stirred at 30 °C for 6 h. The reaction mixture
was quenched with water (20 mL), and extracted with CH2Cl2 (20 mL x 3). The combined organic layer
was dried over sodium sulfate, filtered, concentrated in vacuo and purified by prepꢀHPLC to give 24 mg
(14%) of 1,1,1,3,3,3ꢀhexafluoropropanꢀ2ꢀyl 3ꢀ[1ꢀ(3,4ꢀdifluorophenyl)ꢀ1Hꢀpyrazolꢀ3ꢀyl]azetidineꢀ1ꢀ
1
carboxylate (9) as a white solid. H NMR (400 MHz, CD3OD) δ ppm 8.26 ꢀ 8.14 (m, 1 H), 7.75 (ddd, J =
11.7, 7.1, and 2.6 Hz, 1 H), 7.63 ꢀ 7.52 (m, 1 H), 7.45 ꢀ 7.28 (m, 1 H), 6.51 (d, J = 2.4 Hz, 1 H), 6.16 –
5.99 (m, 1 H), 4.58 ꢀ 4.40 (m, 2 H), 4.34 ꢀ 4.19 (m, 2 H), 4.05 (tt, J = 8.9 and 6.1 Hz, 1 H) LCMS m/z
430.1 [M + H]+
ACS Paragon Plus Environment