Design, Synthesis, and Evaluation of Reversible and Irreversible Monoacylglycerol Lipase Positron Emission Tomography (PET) Tracers Using a "tail Switching" Strategy on a Piperazinyl Azetidine Skeleton

Article abstract of DOI:10.1021/acs.jmedchem.8b01778

Monoacylglycerol lipase (MAGL) is a serine hydrolase that degrades 2-arachidonoylglycerol (2-AG) in the endocannabinoid system (eCB). Selective inhibition of MAGL has emerged as a potential therapeutic approach for the treatment of diverse pathological co

Full text of DOI:10.1021/acs.jmedchem.8b01778

Subscriber access provided by Washington University | Libraries
Article
Design, synthesis and evaluation of reversible and irreversible
monoacylglycerol lipase positron emission tomography (PET) tracers
using a ‘tail switching’ strategy on a piperazinyl azetidine skeleton
Zhen Chen, Wakana Mori, XIAOYUN DENG, Ran Cheng, Daisuke Ogasawara, Genwei Zhang, Michael
Andreas Schafroth, Kenneth Dahl, Hualong Fu, Akiko Hatori, Tuo Shao, Yiding Zhang, Tomoteru Yamasaki,
Xiaofei Zhang, Jian Rong, Qingzhen Yu, Kuan Hu, Masayuki Fujinaga, Lin Xie, Katsushi Kumata, Yuancheng
Gou, Jingjin Chen, Shuyin Gu, Liang Bao, Lu Wang, Thomas Lee Collier, Neil Vasdev, Yihan Shao, Jun-
An Ma, Benjamin F. Cravatt, Christopher Fowler, Lee Josephson, Ming-Rong Zhang, and Steven H Liang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01778 • Publication Date (Web): 04 Mar 2019
Downloaded from http://pubs.acs.org on March 5, 2019
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 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Fujinaga, Masayuki; Department of Radiopharmaceuticals Development,
National Institute of Radiological Sciences, National Institutes for
Quantum and Radiological Science and Technology
Xie, Lin; Kokuritsu Kenkyu Kaihatsu Hojin Hoshasen Igaku Sogo
Kenkyujo, Radiopharmaceuticals Development
Kumata, Katsushi; Kokuritsu Kenkyu Kaihatsu Hojin Hoshasen Igaku
Sogo Kenkyujo, Department of Radiopharmaceuticals Development
Gou, Yuancheng; ChemShuttle, Inc.
Chen, Jingjin; ChemShuttle, Inc.
Gu, Shuyin; ChemShuttle, Inc.
Bao, Liang; ChemShuttle, Inc.
Wang, Lu; Division of Nuclear Medicine and Molecular Imaging,
Massachusetts General Hospital & Department of Radiology, Harvard
Medical School
Collier, Thomas; Division of Nuclear Medicine and Molecular Imaging,
Massachusetts General Hospital & Department of Radiology, Harvard
Medical School
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Vasdev, Neil; Harvard Medical School/Massachusetts General Hospital,
Nuclear Medicine and Molecular Imaging
Shao, Yihan; University of Oklahoma, Chemistry & Biochemistry
Ma, Jun-An; Tianjin University, Department of Chemistry
Cravatt, Benjamin; Scripps Research Institute, Chemical Physiology; The
Scripps Research Institute, Department of Chemical Physiology
Fowler, Christopher; Department of Pharmacology and Clinical
Neuroscience, Umeå University
Josephson, Lee; Division of Nuclear Medicine and Molecular Imaging,
Massachusetts General Hospital & Department of Radiology, Harvard
Medical School
Zhang, Ming-Rong; National Institute of Radiological Sciences,
Department of Radiopharmaceuticals Development
Liang, Steven; Mass General Hospital and Harvard Medical School,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Design, synthesis and evaluation of reversible and irreversible
monoacylglycerol lipase positron emission tomography (PET) tracers
using a ‘tail switching’ strategy on a piperazinyl azetidine skeleton
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Zhen Chen,1,2,# Wakana Mori,3,# Xiaoyun Deng, Ran Cheng, Daisuke Ogasawara, Genwei Zhang,
1
1
4
5
3
4
1
1
3
1
Michael A. Schafroth, Kenneth Dahl, Hualong Fu, Akiko Hatori, Tuo Shao, Yiding Zhang,
3
1
1
1
3
3
Tomoteru Yamasaki, Xiaofei Zhang, Jian Rong, Qingzhen Yu, Kuan Hu, Masayuki Fujinaga, Lin
3
3
6
6
6
6
1
Xie, Katsushi Kumata, Yuancheng Gou, Jingjin Chen, Shuyin Gu, Liang Bao, Lu Wang, Thomas
1
1
5
2
4
7
Lee Collier, Neil Vasdev, Yihan Shao, Jun-An Ma, Benjamin F. Cravatt, Christopher Fowler, Lee
1
3,
1,
Josephson, Ming-Rong Zhang * and Steven H. Liang *
1Division of Nuclear Medicine and Molecular Imaging, Massachusetts General Hospital & Department of
Radiology, Harvard Medical School, Boston, MA, 02114, USA.
2Department of Chemistry, School of Science, Tianjin University, 92 Weijin Road, Nankai District, Tianjin
300072, China
3Department of Radiopharmaceuticals Development, National Institute of Radiological Sciences, National
Institutes for Quantum and Radiological Science and Technology, Chiba, 263-8555, Japan.
4The Skaggs Institute for Chemical Biology and Department of Chemical Physiology, The Scripps Research
Institute, SR107 10550 North Torrey Pines Road, La Jolla, California 92037, USA
5Department of Chemistry and Biochemistry, University of Oklahoma, Norman, Oklahoma 73019, USA.
6ChemShuttle, Inc., 1699 Huishan Blvd., Wuxi, Jiangsu, 214174, China
7Department of Pharmacology and Clinical Neuroscience, Umeå University, SE-901 87 Umeå, Sweden
ACS Paragon Plus Environment

Page 3 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
#These authors contribute equally to this work.
Abstract: Monoacylglycerol lipase (MAGL) is a serine hydrolase that degrades 2-arachidonoylglycerol
2-AG) in the endocannabinoid system (eCB). Selective inhibition of MAGL has emerged as a potential
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
therapeutic approach for the treatment of diverse pathological conditions including chronic pain,
inflammation, cancer and neurodegeneration. Herein we disclose a novel array of reversible and
irreversible MAGL inhibitors by means of tail switching on a piperazinyl azetidine scaffold. We
developed a lead irreversible-binding MAGL inhibitor 8, and reversible-binding compounds 17 and 37
11
18
11
11
which are amenable for radiolabeling with C or F. [ C]8 ([ C]MAGL-2-11) exhibited high brain
1
1
uptake and excellent binding specificity in the brain towards MAGL. Reversible radioligands [ C]17
11
18
18
(
[ C]PAD) and [ F]37 ([ F]MAGL-4-11) also demonstrated excellent in vivo binding specificity
towards MAGL in peripheral organs. This work may pave the way for the development of MAGL-
targeted positron emission tomography (PET) tracers with tunability in reversible and irreversible
binding mechanism.
Keywords: monoacylglycerol lipase; MAGL; 2-AG; positron emission tomography; irreversible;
reversible;
INTRODUCTION
As a lipid signaling network, membrane-bound G-coupled cannabinoid receptors, namely CB and
1
CB , and their endogenous ligands, 2-arachidonoylglycerol (2-AG) and N-arachidonoylethanolamine
2
(
AEA) established the backbone of endocannabinoid (eCB) system.1-5 The eCB system is prominent in
both the central and peripheral nervous system, and its dysfunction has been implicated in a wide range
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of pathological conditionsincluding pain, appetite, inflammation, memory and cognition, and cancer.6-10
Since direct regulation of CB receptors is often accompanied with a series of debilitating adverse
1
effects such as substance abuse, loss of motor and cognition functions,11, 12 recent drug discovery efforts
have been shifted to regulating the levels of AEA or 2-AG. As the most essential endogenous ligands
with endocannabinoid-like activity, AEA and 2-AG are synthesized “on request” in vivo and in the brain
primarily degraded by fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MAGL),
respectively.13-16 In particular, MAGL belongs to serine hydrolase superfamily, which is associated with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
the eCB system as well as eicosanoid and other lipid signaling pathways. In rodents, MAGL is highly
expressed in the CNS as well as several peripheral organs including liver, kidney, adrenal glands and
1
8
brown adipose tissue. In humans, there is a similar CNS MAGL distribution to that in rodents with
high levels of activity in the cerebral cortex, hippocampus and cerebellum, and low levels in the
hypothalamus and pons.19 Considering its prime role in 2-AG hydrolysis in the brain, selective
inhibition of MAGL may represent an alternative and potential therapeutic target for treatment of
diverse pathological conditions including chronic pain, inflammation, cancer and neurodegeneration
2
0-31
without apparent side effects related with direct CB regulation.
1
Positron emission tomography (PET) is a non-invasive and highly sensitive technology in the realm
of molecular imaging, and serves as an ideal tool to quantify biochemical and pharmacological
processes in vivo under normal and disease conditions.32-34 PET studies of MAGL would allow to
achieve in-depth knowledge of MAGL-related pathological changes between normal and disease state,
and in vivo interaction of novel MAGL inhibitors with the target. Development of MAGL PET tracers
would remarkably help to validate promising MAGL inhibitors in clinical trials. As a result, in the past
few years, considerable efforts have been exerted towards this goal despite still with limited success.
ACS Paragon Plus Environment

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The first attempt for PET imaging of MAGL was performed by Hicks et al. with several carbon-11
1
1
11
labeled MAGL inhibitors, including [ C]KML29, and [ C]JJKK-0048. However, all these compounds
35
had limited brain uptake, which impeded their further translation. To date, only three potent MAGL
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
6
11
37-39 11
40
PET tracers, namely [ C]SAR127303,
[ C]MA-PB-1 based on a piperidyl carbamate scaffold,
1
1
41
and [ C]MAGL-0519 based on an azetidinyl oxadiazole scaffold, have been developed to image
MAGL in living brains of rats and non-human primates (NHPs) (Figure 1A). However, most reported
MAGL PET tracers are highly lipophilic (cLogP ca. 3-5), which is often linked with fast metabolic
clearance, poor in vivo stability, and high propensity for off-target promiscuity.42 43 For example, the
2
,5-regioisomer of LY2183240 exhibited poor selectivity between MAGL and FAAH, which could be,
4
3
in some extent, attributed to a high cLogP value of 4.03. Furthermore, the binding mechanism of these
PET tracers was limited to be irreversible and no reversible MAGL PET ligand has been reported to
date. In fact, a reversible MAGL tracer would enable the access to critical quantitative kinetic analysis,
including facilitated measures of binding potential and volume of distribution, for monitoring
neurological response to therapeutics.44, 45 As a result, there is a critical demand for the development of
both irreversible and reversible MAGL PET tracers with favorable lipophilicity and brain kinetics.
A. Previous work:
O
11C
CF3
CF3
O
N
N
O
O
CF3
CF3
Cl
N
CF3
CF3
N
O
*
H
N
O
O
N
S
H311CO
O
O
O
*
CH3
[11
cLogP 4.33
C]SAR127303
[11C]MA-PB-1
cLogP 4.95
[11C]MAGL-0519
cLogP 2.80
B. This work:
Tail =
Tail =
O
11CH3
O
O
Tail
O
11C
O
CF3
CF3
N
S
N
CF3
Cl
18F
N
N
8 ([11C]8)
4
N
49 ([11C]17)
50 ([18F]37)
cLogP 1.87
O
Irreversible (carbamate-type)
Tail Swiching
Reversible (amide-type)
Figure 1. Representative PET tracers for imaging brain MAGL and our work.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
As part of our continuing interest in the development and translation of novel MAGL PET tracers,37,
39, 41
herein we described a novel class of MAGL inhibitors using a ‘tail switching’ strategy,46, 47 wherein
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the ‘tail’ refers to the group that is attached to the unique piperazinyl azetidine skeleton (Figure 1B).48-50
In detail, our medicinal efforts focused on the synthesis of an array of (4-(azetidin-3-yl)piperazin-1-
yl)(thiazol-2-yl)methanone derived carbamates or ureas as irreversible candidate MAGL inhibitors and
(
4-(azetidin-3-yl)piperazin-1-yl)(thiazol-2-yl)methanone derived amides as reversible candidates, with
amenability for radiolabeling with carbon-11 or fluorine-18. Pharmacological studies, molecule
docking and physicochemical evaluations were performed to identify our compound 8 as the most
promising irreversible MAGL inhibitor, and compounds 17 & 37 as the most promising reversible
MAGL inhibitors, worthy of radiolabeling and in vivo PET translational studies. With innovative and
1
1
18
efficient C- and F-labeling strategies, we evaluated the brain permeability, binding specificity and
11
11
11
kinetics of these lead radioligands 48 ([ C]8), 49 ([ C]17) and 50 ([ C]37) by PET experiments in
rodents. While irreversible MAGL tracer 48 demonstrated excellent in vitro potency and selectivity, in
vivo binding specificity and stability in the brain, our reversible MAGL tracers 49 and 50 demonstrated
high-level specific binding to MAGL in a peripherally-restricted manner. As proof-of-concept, we were
able to unveil the underlying cause of low brain accumulation for 49 and 50, the most potent reversible
MAGL tracers in our design, thus paving the way for future development of reversible MAGL PET
tracers.
RESULTS AND DISCUSSION
ACS Paragon Plus Environment

Page 7 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Chemistry. A focused library of (4-(azetidin-3-yl)piperazin-1-yl)(thiazol-2-yl)methanone derived
carbamates or ureas 8-13 as irreversible MAGL inhibitor candidates and (4-(azetidin-3-yl)piperazin-1-
yl)(thiazol-2-yl)methanone derived amides 16-22, 36-38 as reversible MAGL inhibitor candidates were
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
18
synthesized, the scaffold of which are amenable for C- or F-labeling. As summarized in Scheme 1,
the S 2 displacement reaction between tert-butyloxycarbonyl (Boc)-protected piperazine 1 and 1-
N
benzhydrylazetidin-3-yl methanesulfonate 2 readily proceeded to give 3 in 85% yield. Trifluoroacetic
acid (TFA)-induced deprotection of the Boc group from 3 led to isolation of 4 in nearly quantitative
yield, which subsequently coupled with thiazole-2-carboxylic acid 5 to produce 6 in high efficiency in
the presence of 1-hydroxybenzotriazole hydrate (HOBT) and N-(3-dimethylaminopropyl)-N’-
ethylcarbodiimide hydrochloride (EDC•HCl). 1-Chloroethyl chloroformate triggered deprotection of
the diphenylmethyl group from 6 released azetidine 7 in 86% yield, which served as a crucial precursor
1
1
for the following C-labeling. To synthesize irreversible MAGL inhibitor candidates 8-13, we deployed
several parallel approaches for the introduction of different carbonyl-R groups. The combination of
1
,1,1,3,3,3-hexafluoro-2-propanol, 4-nitrophenyl chloroformate, 4-dimethylaminopyridine (DMAP),
pyridine and Et N in CH Cl proved optimal to generate 8 in 20% yield. For candidates 9-12,
3
2
2
triphosgene was found to be a superior activating reagent, thus producing the corresponding carbamates
and 12, as well as triazolyl carbonyls 10 and 11 in 7-43% yields. Candidate 13 was isolated in 13%
9
yield by the treatment of azetidine 7 with N,N'-disuccinimidyl carbonate.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ph
Ph
N
S
Ph
N
Ph
Ph
OH
H
Ph
N
N
(i)
(ii)
+
N
2
5
O
N
N
Boc
N
MsO
(iii)
N
N
1
3
HN
4
Boc
Ph
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NHHCl
N
Ph
(iv)
N
R
S
N
S
(v) or (vi) or (vii)
S
N
N
N
N
N
N
N
O
O
O
8-13
6
7
Entry
8
9
10
11
12
13
O
O
O
CF3
N
N
N
O
CF3
N
R
N
N
CF3
O
O
N
N
O
O
o
Scheme 1. Synthesis of irreversible MAGL inhibitors 8-13. Conditions: (i) DIPEA, MeCN, 80 C for 12 h; 85% yield; (ii)
TFA, CH
2
Cl
2
3
, rt, 12 h; 99% yield; (iii) HOBT, EDC•HCl, Et N, DMF, rt, 12 h; 78% yield; (iv) 1-chloroethyl chloroformate,
o
CH
2
Cl
2
, rt, 2 h; then MeOH, 35 C 2 h; 86% yield; (v) 1,1,1,3,3,3-hexafluoro-2-propanol, 4-nitrophenyl chloroformate, DMAP,
3 2 2
pyridine, Et N, CH Cl , rt, 5 h; 20% yield for 8; (vi) 2,2,2-trifluoroethanol (for 9) or 1,2,4-triazole (for 10) or 1H-
3 2 2
benzo[d][1,2,3]triazole (for 11) or 2-hydroxyisoindoline-1,3-dione (for 12), triphosgene, DMAP, Et N, CH Cl , rt, 4 h; 43%
3 2 2
yield for 9; 28% yield for 10; 7% yield for 11; 29% yield for 12; (vii) N,N'-disuccinimidyl carbonate, Et N, CH Cl , rt,
overnight; 14% yield for 13; DIPEA = N,N-diisopropylethylamine; TFA = trifluoroacetic acid; HOBT = 1-
hydroxybenzotriazole hydrate; EDC•HCl = N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride; Et
3
N =
triethylamine; DMF = N,N-dimethylformamide; DMAP = 4-dimethylaminopyridine.
To synthesize reversible candidate MAGL inhibitors 16-22 and 36-38, methyl 4-bromo-3-
methoxybenzoate was used as the starting material (Scheme 2A). Cross-coupling reactions with aryl
boronic acid followed by LiOH-mediated hydrolysis readily provided [1,1'-biaryl]-4-carboxylic acid 15
in 32-89% yield over two steps. For Suzuki-type reactions with phenyl boronic acids, Pd(PPh ) proved
3
to be the optimal catalyst, while in the case of heteroaryl boronic acids as coupling partners, PdCl (dppf)
2
ACS Paragon Plus Environment

Page 9 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was utilized to provide good yields. Subsequent coupling reactions of 15 with piperazinyl azetidine 7
in the presence of HOBT and EDC•HCl proceeded smoothly to give amide-type MAGL inhibitors 16-
2
2 in 21-40% yields. For the synthesis of 36-38, ethyl 4-fluorobenzoate was utilized as the starting
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
material (Scheme 2B). Amination reactions with hydroxyl azetidine, hydroxyl pyrazole or hydroxyl
piperidine occurred smoothly under basic conditions, thus delivering 24-26 in moderate to high
efficiencies (38-82% yields). Activation of the hydroxyl group by treatment with methane sulfonyl
chloride afforded the mesylate compounds 27-29 in excellent yields (80-84%), which were readily
fluorinated by tetrabutylammonium fluoride (TBAF) to generate 30-32 in 23-38% yields. The poor
yields observed in the fluorination reactions can likely be attributed to the propensity of β-elimination
of methanesulfonate derivatives 27-29 under basic conditions. Ultimately, candidates 36-38 were
obtained in 18-28% yield over two steps, namely LiOH-promoted hydrolysis of 30-32 followed by
condensation with piperazinyl azetidine 7.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A. Synthesis of reversible MAGL inhibitor candidates 16-22
O
O
O
OMe
Ar
OMe
Br
Ar B(OH)2
(i) or (ii)
OMe
Ar
7
N
MeO
HO
S
N
(
iv)
(iii)
N
14
15
N
1
6-22
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Entry
Ar
16
17
18
19
20
21
22
CF3
N
N
O
S
N
Cl
N
B. Synthesis of reversible MAGL inhibitor candidates 36-38
O
O
O
EtO
(vi)
EtO
(vii)
(v)
EtO
m
n
m
N
OH
N
OMs
F
27, m = 1, n = 0
8, m = 1, n = 1
9, m = 2, n = 1
n
23
24, m = 1, n = 0
25, m = 1, n = 1
2
2
2
6, m = 2, n = 1
O
O
O
EtO
N
(viii)
HO
7
(iv)
m
n
m
N
F
m
n
S
N
N
F
N
F
30, m = 1, n = 0
31, m = 1, n = 1
32, m = 2, n = 1
N
N
n
33, m = 1, n = 0
34, m = 1, n = 1
3
3
3
6, m = 1, n = 0
7, m = 1, n = 1
8, m = 2, n = 1
O
35, m = 2, n = 1
o
3 4 2 3 2
Scheme 2. Synthesis of reversible MAGL inhibitors 16-22 and 36-38. Conditions: (i) Pd(PPh ) , K CO , toluene/H O, 100 C,
o
overnight; (ii) PdCl
2
(dppf), K
2
CO
3
, 1,4-dioxane/H
2
O, 105 C, overnight; (iii) LiOH, THF/H
2
O, rt, overnight; (iv) HOBT,
3
EDC•HCl, Et N, DMF, rt, 12 h; 22% yield for 16; 25% yield for 17; 21% yield for 18; 23% yield for 19; 24% yield for 20;
3
2
2
9% yield for 21; 40% yield for 22; 32% yield for 36; 30% yield for 37; 19% yield for 38; (v) azetidin-3-ol hydrochloride (for
o
o
2 3
4) or pyrrolidin-3-ol hydrochloride (for 25) or piperidin-4-ol (for 26), K CO , DMSO, 180 C for 2 h (for 24) or 120 C for
3 2 2
4 h (for 25 and 26); 38% yield for 24; 79% yield for 25; 82% yield for 26; (vi) MsCl, Et N, CH Cl , rt, overnight; 81% yield
o
for 27; 84% yield for 28; 80% yield for 29; (vii) TBAF, THF, 70 C, 2 h; 23% yield for 30; 38% yield for 31; 26% yield for
o
32; (viii) LiOH, THF/MeOH/H
2
O, 40 C, 16 h; 73% yield for 33; 92% yield for 34; 95% yield for 35; dppf = 1,1′-
bis(diphenylphosphino)ferrocene; HOBT = 1-hydroxybenzotriazole hydrate; EDC•HCl = N-(3-Dimethylaminopropyl)-N′-
ethylcarbodiimide hydrochloride; DMF = N,N-dimethylformamide; DMSO = methyl sulfoxide; MsCl = methanesulfonyl
chloride; TBAF = tetrabutylammonium fluoride.
ACS Paragon Plus Environment

Page 11 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Pharmacology. Compounds 8-13, 16-22 and 36-38 were investigated for their potency and
selectivity towards MAGL in vitro. For irreversible candidates 8-13, we determined their in vitro ability
3
3
to inhibit MAGL hydrolysis of [ H]2-oleoylglycerol ([ H]2-OG), a tritiated 2-AG analog, according to
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
1
our previously reported protocol. As outlined in Table 1, Figure 2A and Figure S1, candidate 8
containing a hexafluoroisopropanol leaving group demonstrated the most promising potency towards
inhibition of MAGL activity with an IC value to be 0.88 nM, whereas all of the other candidates 9-13
5
0
indicated inferior potency (Figure S1). The residual activity seen in Figure 2B reflects the fact that
whilst is the predominant hydrolytic enzyme in the brain, it is not the only one: Blankman et al.15
reported that MAGL was responsible for ~85% of the hydrolysis of 2-AG in the brain, and the present
data is consistent with this report. Inhibitor 8 exhibited preincubation-time-dependent inhibition at three
tested concentrations (0.3, 1 and 3 nM, Figure 2B), which is in line with the irreversible binding
mechanism. Although compound 10 possessing a triazole leaving group also exhibited good potency
(
IC =10 nM), no further evaluation was conducted considering the poor blood-brain barrier (BBB)
50
penetration ability for an irreversible MAGL PET tracer with a triazole leaving group.39
Entry
8
9
10
11
12
13
MAGL-0519
8.4 ± 0.2
IC50 (nM) 0.88 ± 0.05 > 1000 10.0 ± 4.2 87.1 ± 12.4 >1000 27.5 ± 8.1
Table 1. IC50 values of compounds 8-13 for inhibition of MAGL activity with MAGL-0519, a known irreversible MAGL
inhibitor, as reference.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
Figure 2. (A) Concentration-response curses of 8 for inhibition of [ H]2-OG hydrolysis by rat brain MAGL (1 h preincubation
3
between enzyme source and inhibitor prior to addition of substrate); (B) Time-dependent inhibition of [ H]2-OG hydrolysis at
three different concentrations (0.3, 1 and 3 nM).
For reversible candidates 16-22 and 36-38, we initially evaluated their potency towards MAGL at
a single concentration of 1 µM utilizing two different assays: (1) activity-based protein profiling (ABPP)
3
studies and (2) hydrolysis of [ H]2-OG. As depicted in Figure 3A and S2, both methods demonstrated
compounds 17, 37 and 38 as the most potent MAGL inhibitors. The IC values of candidates 17, 37
5
0
and 38 were further determined as 2.7 nM, 11.7 nM and 15.0 nM, respectively, via ABPP assays (Figure
B). As proof-of-concept, the reversibility of binding mechanism for candidates 17 and 37 was further
3
investigated. In the case of 17, we utilized an ABPP assay (Figure 3C). In this case, enzyme activity
was measured by treatment with FP-rhodamine. For a reversible inhibitor, the compound would
compete with FP-rhodamine, and each dissociation from the enzyme would present a new target for FP-
rhodamine. Therefore, for reversible inhibitors, the levels of MAGL labeled with FP-rhodamine would
increase over time. This was shown to be the case for 17, but not for the irreversible inhibitor MJN110.
ACS Paragon Plus Environment

Page 13 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
In the case of 37, we investigated the mode of inhibition using [ H]2-OG as substrate (Figures 3D, E).
There was no preincubation time-dependent increase in observed inhibition of the 2-OG hydrolytic
activity of the samples, which would have been expected for an irreversible inhibitor (Figure 3D). Jump
dilution experiments, whereby samples are preincubated with inhibitor and then diluted prior to addition
of substrate suggested no recovery of inhibition following a 20-fold dilution (Figure 3E). However,
this behavior can be found under certain conditions (Kiapp/[E]≥10, short incubation times with substrate)
for tight-binding reversible inhibitors when the high potency is due to longer residence times, i.e. slower
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
52
11
rates of dissociation. Based on these preliminary results, we selected 17 for C-labeling, and 37 for
18F-labeling, as lead reversible inhibitors for further in vivo evaluation.
Figure 3. Inhibition properties of MAGL activity by compounds 18-22 and 36-38. (A) Rat brain MAGL activity in the presence
of compounds 16-22 and 36-38 at 1 μM concentration determined by ABPP assay; (B) Concentration-response curves of 17
and 37-38 for inhibition of rat brain MAGL activity determined by ABPP assay with pristimerin, a known reversible MAGL
inhibitor, as reference (Figure S3 in the Supporting Information); (C) MAGL activity determined at different incubation time
points with 17 and the irreversible inhibitor MJN110 as control. Recovered MAGL activity over time indicates the binding of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
compound 17 is reversible while the irreversible inhibitor MJN110 shows no significant time-dependent changes. (D) Time-
3
dependent inhibition of [ H]2-OG hydrolysis at three different concentrations of compound 37 (30, 100 and 300 nM). (E)
3
Inhibition of 0.5 µM [ H]2-OG hydrolysis by compound 37 without preincubation (red and black bars) or with 60 min of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
preincubation followed by dilution to reduce the free concentration 20-fold (gray bar). Given that this compound showed no
3
time-dependence towards inhibition of [ H]2-OG hydrolysis (data not shown), the data are consistent with a tight-binding but
reversible inhibitor. All data are mean ± SD, n = 3-5. *p < 0.05, **p ≤0.01, and ***p ≤0.001.
As shown in Table 2, the selectivity of these three lead compounds 8, 17 and 37 towards MAGL
over FAAH was further investigated in ABPP assays, and none of them showed significant inhibition
towards FAAH at a concentration up to 10 µM. In addition, for all three lead compounds, no significant
interactions were observed with CB and CB receptors (up to a concentration of 30 μM, Figure 4) as
1
2
well as ABHD6 and ABHD12 (up to a concentration of 10 μM), both of which also belong to serine
hydrolase superfamily and play a vital role in the metabolism of 2-AG in the brain.
Lipophilicity of candidate compounds is an essential factor for the prediction of BBB permeability
with 1.0 to 3.5 as the favorable range.53-55 By means of the ‘shake flask method’, namely liquid-liquid
5
6
partition between PBS and n-octanol, the LogD values of 8, 17 and 37 were 1.90 ± 0.38, 3.35 ± 0.50
and 2.70 ± 0.04, respectively (n = 3) (Table 1). The topological polar surface areas (tPSA) for these
compounds were determined by in silico prediction with ChemBioDraw Ultra 14.0, and all of them
exhibited reasonable tPSA (Table 2), which, together with the lipophilicity results, indicated a high
possibility of sufficient brain permeability.
ACS Paragon Plus Environment

Page 15 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
FAAHa
ABHD6a
ABHD12a
CB1b
CB2b
logDc
tPSAd
Entry
8
<
30% activity at 1
and 10 M of 8, 17
and 37
< 30% activity at 1 and 10
M of 8, 17 and 37 in
ABHD6/12 ABPP assay
< 30% activity at 30 M of 8,
17 and 37 CB /CB2 agonist
and antagonist assay
1.90 ± 0.38
3.35 ± 0.50
65.45
65.45
1
7
7
1
3
59.46
2
.70 ± 0.04
a
Table 2. Binding potency and physicochemical properties of compounds 8, 17 and 37. Determined by ABPP assay on rat
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
b
brain membrane (MJN110 as positive control); Determined by CB
1
/CB
2
agonist and antagonist assays (CP55940, rimonabant,
1 2 1 2
and SR144528 as positive control in CB /CB agonist assays, CB antagonist assay, and CB antagonist assay, respectively);
c
d
Measured by “shake-flask” method quantified by LC-MS/MS. Topological polar surface area (tPSA) was calculated by
ChemBioDraw Ultra 14.0.
1 2
Figure 4. Pharmacological interactions of compounds 8, 17 and 37 with CB /CB receptors. (A) Agonism of 8, 17 and 37 in
1 1
the CB assay with CP55940 as control; (B) Antagonism of 8, 17 and 37 in the CB assay with rimonabant as control; (C)
2 2
Agonism of 8, 17 and 37 in the CB assay with CP55940 as control and (D) Antagonism of 8, 17 and 37 in the CB assay with
SR144528 as control. All data are mean ± SD, n = 3-5.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Molecular docking studies. Based on the fact of excellent binding affinities for lead compounds
, 17 and 37 (0.88, 2.7, 11.7 nM, respectively), we then performed molecular docking studies with the
8
purpose of identifying their possible interactions with MAGL and the corresponding binding domain
Figure 5). A 1.35-Å resolution crystal structure of a soluble form human monoglyceride lipase (MGLL)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
was downloaded (PDB ID: 3PE6) and its original ligand57 was removed before performing the
molecular docking studies. First, we reproduced the binding pose of the original ligand within 3PE6
using Autodock Vina, which yielded an excellent overlapping between the docked and original poses
(
Figure S4, Supporting Information). Subsequently, selected MAGL inhibitors 8, 17 and 37 were
docked onto the 3PE6 structure (Figure 5 and Figure S5, Supporting Information). As shown in Figure
A and 5B, both compounds 8 and 17 entered the 2-AG binding pocket and were located near the
5
catalytic triad site, formed by Ser122, His269 and Asp239. Since the catalytic triad is essential for
catalyzing 2-AG hydrolysis, compounds 8 and 17 reduced the accessibility of the active site to 2-AG,
therefore inhibiting the 2-AG hydrolysis. Meanwhile, compound 37 was only partially docked into the
2
-AG binding pocket, which blocked the “lid” of the pocket and prevented the 2-AG substrate from
accessing the catalytic site (Figure 5C). It is also worth mentioning that although the Vina docking
studies suggested that the docked pose of 37 did not bind to the oxyanionic hole or 37 was situated in a
different binding pocket, more in-depth computational studies, including molecular dynamic simulation,
are necessary to verify this hypothesis. Inside the binding pocket, we found π-π stacking interactions
and hydrogen bonds. The π-π stackings were observed between the phenyl group of compounds 17, 37
and nearby Phe159 side chains (Figures 5B and 5C). And the Hydrogen bonds were found between
compound 8 and Ala51 (Figure 5A), and also between compound 17 and nearby Ala51 and Met123
ACS Paragon Plus Environment

Page 17 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(Figure 5B). Overall, these docking results provided supportive evidence for the inhibitory mechanism
of aforementioned compounds.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Molecular docking structures of compounds 8 (yellow in A; pre-covalent docking state), 17 (blue in B) and 37
(magenta in C) onto MAGL. The catalytic triad was labeled as Asp239, His269 and Ser122. Hydrogen-bonds between
compounds and the MAGL protein were connected with solid lines. The bottom-right insets at each panel showed the docking
pose of each compound into (or at the opening end of) the 2-AG binding pocket. The PDB ID of the protein structure is 3PE6.
Radiochemistry. The encouraging results of pharmacology and physiochemical property and
docking studies prompted us to perform radiolabeling of 8, 17 and 37 with carbon-11 or fluorine-18. In
the case of 37, we used its racemic form for preliminary proof-of-concept studies for brain permeability
and binding specificity. The significance of labeling site selection in irreversible serine hydrolase PET
tracers has been demonstrated by Wilson58, 59 and our group.39, 41 Therefore, for irreversible lead
1
1
11
11
compound 8, we utilized two C-carbonylation approaches with [ C]COCl and [ C]CO , respectively
2
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
(Scheme 4A). The piperazinyl azetidine precursor 7 for C-carbonylation was synthesized as per the
1
1
11
method depicted in Scheme 1. Initial radiosynthesis of 48 ([ C]8) was performed with [ C]COCl
2
generated from our previously reported procedure.39 Reaction of [ C]COCl with 1,1,1,3,3,3-
11
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hexafluoro-2-propanol (HFIP) facilitated by 1,2,2,6,6-pentamethylpiperidine (PMP) followed by
1
1
addition of azetidine 7 smoothly occurred to deliver C-labeled carbamate 48 in an average 18% (n =
1
1
6) radiochemical yield (RCY) relative to starting [ C]CO (decay corrected). The radiochemical purity
2
was ≥99% and the molar activity was 30–52 GBq/µmol (0.81–1.40 Ci/µmol). Alternative
11
radiosynthesis of 48 was carried out based on our previously reported ‘in-loop’ [ C]CO fixation
2
6
0
11
method. Treatment of [ C]CO with 2-tert-butylimino-2-diethylamino-1,3-dimethylperhydro-1,3,2-
2
diazaphosphorine (BEMP) and piperazinyl azetidine 7, followed by successive addition of HFIP and
phosphorus(V) oxychloride (POCl ) in acetonitrile (MeCN), readily produced 48 in an average 2.5%
3
1
1
decay-corrected RCY based on starting [ C]CO at the end-of-synthesis. The radiochemical purity was
2
greater than 99% (n = 2) and the molar activity was as high as 288.6 GBq/µmol (7.8 Ci/µmol).
For lead compounds 17 and 37 with a reversible binding mechanism, we deployed two distinct
1
1
18
labeling methods, namely C-methylation of phenolic precursor 44 for lead compound 17 and F-
fluorination of mesylate precursor 47 for lead compound 37, respectively (Scheme 4B and 4C). The
general procedure for preparation of precursors 44 and 47 was outlined in Scheme 3A and 3B,
respectively. Specifically, synthesis of phenolic precursor 44 commenced with the protection of phenol
derivative 39 with chloromethyl methyl ether (MOMCl) to give bromobenzene derivative 40 in 90%
yield. Subsequently, Pd(PPh ) -catalyzed cross-coupling reaction of 40 with (4-chloro-3-
3
4
(
trifluoromethyl)phenyl)boronic acid 41, followed by LiOH-triggered hydrolysis of the ester group,
readily proceeded to give [1,1'-biphenyl]-4-carboxylic acid 42 in 89% yield over two steps. Coupling
ACS Paragon Plus Environment

Page 19 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 42 with piperazinyl azetidine 7 in the presence of HOBT and EDC occurred to deliver azetidine
amide 43 in 24% yield. The following deprotection of the methoxylmethyl ether (MOM) group from
4
3 was achieved with 6 N HCl, thus affording the desired phenolic precursor 44 in 80% yield. On the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
other hand, the synthetic route to mesylate precursor 47 involved a K CO promoted displacement of
2
3
ethyl 4-fluorobenzoate 23 by pyrrolidin-3-ol hydrochloride, followed by LiOH-mediated hydrolysis of
the ester group, thus producing 4-(3-hydroxypyrrolidin-1-yl)benzoic acid 45 in 77% overall yield.
Coupling of 45 with piperazinyl azetidine 7, EDC and HOBT led to the isolation of hydroxyl-containing
azetidine amide 46, which was subsequently protected with methanesulfonyl chloride (MsCl) to
generate the desired mesylate precursor 47 in 16% yield over two steps. As shown in Scheme 4B, the
1
1
radiotracer 49 ([ C]17) with reversible binding mechanism was prepared by reacting the phenolic
1
1
11
precursor 44 with [ C]CH I in 25% decay-corrected RCY based on starting [ C]CO at the end of
3
2
1
8
18
synthesis with molar activity greater than 97.5 GBq/µmol. The F-labeled radiotracer 50 ([ F]37) was
18
obtained by reacting the mesylate precursor 47 with [ F]fluoride in the presence of K CO and
2
3
4
,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8,8,8]hexacosane (Kryptofix222) in an average RCY of 39%
(decay corrected) with high radiochemical purity (>99%) and excellent molar activity (>110
GBq/µmol). All these radiotracers exhibited excellent in vitro stability without radiolysis up to 90 min
after formulation. The efficient radiosynthesis, as well as excellent radiochemical purity and molar
activity of 48-50 paved the way for subsequent in vivo investigation.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A. Synthesis of precursor 44 for 11C-labelling of 17
(
HO) B
CF3
Cl
2
O
O
O
OMOM
(i)
41
HO
7
OH
Br
OMOM
Br
MeO
MeO
CF3
Cl
(
ii) (iii)
(iv)
39
42
N
40
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
OH
OMOM
N
(v)
CF3
CF3
Cl
S
N
S
N
N
N
N
N
Cl
O
O
43
44
B. Synthesis of precursor 47 for 18F-labelling of 37
O
O
O
(vi)
(vii)
7
EtO
HO
EtO
(
iv)
N
OH
N
OH
F
4
5
23
25
O
O
N
N
(viii)
S
N
S
N
N
OH
N
OMs
N
N
N
N
46
47
O
O
2 3
Scheme 3. Synthesis of labeling precursors 44 and 47 for MAGL PET tracers. Conditions: (i) MOMCl, K CO , acetone, rt,
overnight; 95% yield; (ii) Pd (PPh
3
)
4
, K
2
CO
3
, toluene/H
2
O, reflux, overnight; then (iii) LiOH, THF/H
2
O, rt, overnight; 90%
3
yield over two steps; (iv) HOBT, EDC•HCl, Et N, DMF, rt, 12 h; 24% yield for 43; 22% yield for 46; (v) 6 N HCl, MeOH, 66
oC, 2 h; 80% yield; (vi) pyrrolidin-3-ol hydrochloride, K
o
2
3 2
CO , DMSO, 120 C, 24 h; 79% yield; (vii) LiOH, THF/MeOH/H O,
o
4
3 2 2
0 C, overnight; 97% yield; (viii) MsCl, Et N, CH Cl , rt, overnight; 70% yield; HOBT = 1-hydroxybenzotriazole hydrate;
EDC•HCl = N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride; DMF = N,N-dimethylformamide; DMSO =
methyl sulfoxide; MsCl = methanesulfonyl chloride.
ACS Paragon Plus Environment

Page 21 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A. Synthesis of irrversible tracer 48 with [11C]COCl2 or [11C]CO2
S
O
S
NHHCl
(i) or (ii)
11C
CF3
N
N
N
N
N
N
N
O
O
7
CF3
O
4
8 ([11C]8)
B. Synthesis of reversible tracer 49 with [11C]MeI
O
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
11CH3
O
OH
O
N
N
CF3
(iii)
CF3
Cl
S
N
S
N
N
N
N
N
Cl
44
O
O
S
9 ([11C]17)
4
C. Synthesis of reversible tracer 50 with [18F]TBAF
O
N
O
N
(iv)
S
N
N
N
18F
N
OMs
N
N
N
47
N
50 ([18F]37)
O
O
1
1
o
Scheme 4. Radiosynthesis of tracers 48-50. Conditions: (i) [ C]COCl2, HFIP, PMP, THF; 30 C, 1 min. 18% decay-corrected
11
11
2 3
RCY (ii) [ C]CO , BEMP, HFIP, POCl , rt, 1 min. 2.5% decay-corrected RCY; (iii) [ C]MeI, NaOH, DMF, 70 °C, 5 min,
1
8
2
5% decay-corrected RCY; (iv) [ F]KF, DMSO, 120 °C, 10 min, 39% decay-corrected RCY. HFIP = 1,1,1,3,3,3-hexafluoro-
2
-propanol; PMP = 1,2,2,6,6-pentamethylpiperidine; BEMP = 2-tert-butylimino-2-diethylamino-1,3-dimethylperhydro-1,3,2-
diazaphosphorine.
1
1
Preliminary PET imaging studies of irreversible MAGL tracer 48 ([ C]8). Dynamic PET
acquisitions of radioligand 48 were carried out in Sprague-Dawley rats for 90 mins. Representative PET
images in the brain (sagittal and coronal, summed images 0-90 min) and time-activity curves are shown
in Figure 6 and Figure S6 in the Supporting Information. Compound 48 demonstrated excellent BBB
penetration with maximum standard uptake value (SUV) of 2.26 at 3.5 min in the whole brain, and
heterogeneous distribution with high uptake in cerebellum, striatum and hippocampus and low uptake
in pons (Figure 6A and Figure S6, Supporting Information). No significant washout (ratio of SUV3.5
/
SUV90 min = 1.2) was observed during PET scans, which was aligned with characteristic plateaued
min
time−activity curves of typically irreversible binding PET tracers.40, 58, 61 Pretreatment with KML29 (3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
mg/kg), a widely-used MAGL inhibitor, remarkably decreased the uptake of 48 in various brain regions
and the whole brain uptake (~50% reduction based on area under curve, AUC), which demonstrated
excellent in vivo specificity of 48 towards MAGL (Figure 6B and Figure S6). We then investigated
the uptake, biodistribution and clearance of 48 by whole body distribution in mice at five time points
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(1, 5, 15, 30 and 60 min) after tracer injection (Figure 6C and Table S1, Supporting Information).
Several organs including liver, adrenal glands, kidney and brown adipose tissue (BAT) exhibited high
radioactivity levels at 30 min (>5% ID/g, injected dose per gram of wet tissue), which is consistent with
62
the expression of MAGL in rodents. The radioactivity levels in heart and lungs reached a plateau at 5
min, followed by slow washout. These results together with high activity remaining in the kidneys,
liver, and small intestine suggested urinary and hepatobiliary elimination of 48. To investigate the in
vivo stability of 48, we carried out radiometabolic analysis at 30 min post-injection of the radiotracers
in mouse brain homogenate for 48. The fraction of remaining parent 48 was determined to be 96% by
radioHPLC, indicating excellent in vivo stability in the brain.
ACS Paragon Plus Environment

Page 23 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Representative PET images and time-activity curves of 48 in rat brain: (A) Baseline; (B) Blocking studies with
KML29 (3 mg/kg); (C) Whole body ex vivo biodistribution studies. The brain uptake was measured and quantified by the
standardized uptake value (SUV). Asterisks indicate statistical significance. p < 0.05, p ≤ 0.01, and *** p ≤ 0.001. AGs =
*
**
adrenal glands; SI = small intestine; WAT = white adipose tissue; BAT = brown adipose tissue.
11
18
Preliminary PET imaging studies of reversible MAGL tracers 49 ([ C]17) and 50 ([ F]37).
Dynamic PET acquisitions of reversible radioligands 49 and 50 were also performed in Sprague-
Dawley rats for 90 mins, and both exhibited insufficient BBB permeability in rats (ca. 0.4 and 0.3 SUV,
respectively, Figure 7A-B and 8A-B). High levels of radioactivity were observed in brown adipose
tissue (BAT), kidneys and liver (Figure 7A and 8A), which is consistent with the expression of MAGL
in periphery and confirmed by subsequent ex vivo biodistribution studies (Figure 7C, 8C and Table
S2, S3, Supporting Information).62 In contrast to the irreversible radioligand 48, both 49 and 50
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
exhibited initial high uptake (> 15% ID/g at 1 min) in heart and lungs followed by fast clearance (ratio
of %ID/g 1 min/60 min in heart: 3.49 for 49, 6.34 for 50; ratio of %ID/g1 min/60 min in lungs: 6.80 for 49, 10.38
for 50). Besides, the radioactivity of 49 and 50 were also rapidly washed out from blood with the SUV1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
/
SUV60 min ratios to be 5.1 and 3.5, respectively. These results together with high bound activity in
min
the kidneys, liver, and small intestine suggested fast urinary and hepatobiliary elimination of 49 and 50.
To test the in vivo specificity of 49 and 50 towards MAGL, we carried out blocking experiments by
pre-treatment of KML29 (3 mg/kg). While 50 exhibited significant blocking effect in two MAGL-rich
organs including adrenal glands (80%) and BAT (88%), 49 demonstrated comparable result in BAT (63%
decrease), and no statistical significance was observed for the uptake in adrenal glands (13% decrease)
between the baseline and blocking conditions (Figure 7D and 8D). The increased level of non-specific
binding of compound 49 in adrenal glands and BAT was likely attributed to its higher lipophilicity. In
tissues with low MAGL expression, such as lungs and muscle, no obvious contrast was observed for
both 49 and 50 between baseline and blocking conditions, indicating a limited window to show specific
binding in MAGL-deficient organs. Furthermore, in MAGL-deficient organs such as lungs and muscle,
no significant changes of uptake were observed for both 49 and 50. The peripherally-restricted
distribution of 49 and 50 may be attributed to their interaction with P-glycoprotein (Pgp) and breast
cancer resistance protein (Bcrp), both of which are crucial components of ATP-binding cassette (ABC)
efflux transporter family co-localizing at the BBB. To verify this hypothesis, we conducted PET
imaging studies of 49 and 50 in wild-type and Pgp/Bcrp knockout (ABCB1a/1b-/-ABCG2-/-) mice. As
outlined in Figure 7E-F and 8E-F, the brain uptake was significantly increased in Pgp/Bcrp knockout
mice than that of wide-type mice, and reached a plateau of 1.4 SUV at 2.5 min for 49 and 1.0 SUV at
1
1 min for 50, respectively. These results demonstrated that 49 and 50 had intensive interactions with
ACS Paragon Plus Environment

Page 25 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABC efflux transporters at the BBB including Pgp and Bcrp, which may explain their limited brain
accumulation at tracer dose. The extensive interactions of 49 and 50 with ABC efflux transporters are
probably attributed to the nature that they are nitrogenous compounds and characterized by one or more
positive charges in the physiological conditions. To investigate the in vivo stability of 49 and 50, we
performed radiometabolic analysis at 30 min post-injection of the radiotracers in mouse plasma
homogenate. The fractions of parent 49 and 50 were determined by radioHPLC to be 62% and 45%,
respectively, indicating reasonable metabolic stability.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Representative PET images and time-activity curves of 49: (A) Baseline PET images in SD rat (summed 0-90 min);
(B) Baseline PET images (summed 0-90 min) and time-activity curve of 49 in rat brain; (C) Whole body ex vivo biodistribution
in mice at five different time points (1, 5, 15, 30 and 60 min) post injection of 49. All data are mean ± SD, n = 4. (D) ex vivo
Specificity measured by the uptake of 49 in MAGL-rich tissues adrenal glands and brown adipose tissue (baseline and blocking
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with KML29 (3 mg/kg)); (E) Representative PET images of wild-type and Pgp/Bcrp knockout mouse (summed 0-90 min); (F)
Representative PET images and time-activity curve of 49 of wild-type and Pgp/Bcrp knockout mouse (summed 0-90 min).
Asterisks indicate statistical significance. p < 0.05, p ≤0.01, and *** p ≤0.001. AGs = adrenal glands; SI = small intestine;
*
**
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
WAT = white adipose tissue; BAT = brown adipose tissue.
Figure 8. Representative PET images and time-activity curves of 50. (A) Baseline PET images in SD rat (summed 0-90 min);
(B) Baseline PET images (summed 0-90 min) and time-activity curve of 50 in rat brain; (C) Whole body ex vivo biodistribution
in mice at five different time points (1, 5, 15, 30 and 60 min) post injection of 50. All data are mean ± SD, n = 4; (D) ex vivo
Specificity measured by the uptake of 50 in MAGL-rich tissues including adrenal glands (AGs) and brown adipose tissue
(BAT) (baseline and blocking with KML29 (3 mg/kg)); (E) Representative PET images of wild-type and Pgp/Bcrp knockout
mouse (summed 0-90 min); (F) Representative PET images and time-activity curve of 50 of wild-type and Pgp/Bcrp knockout
ACS Paragon Plus Environment

Page 27 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mouse (summed 0-90 min). Asterisks indicate statistical significance. p < 0.05, p ≤ 0.01, and *** p ≤ 0.001. AGs = adrenal
*
**
glands; SI = small intestine; WAT = white adipose tissue; BAT = brown adipose tissue.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CONCLUSION
We have established a focused library of MAGL inhibitors with irreversible or reversible binding
mechanisms by the utilization of a tail switching strategy on a piperazinyl azetidine scaffold. Potency,
selectivity as well as irreversible or reversible binding profiles were determined in vitro by two different
3
tissue-based assays, namely [ H]2-OG hydrolysis and ABPP. Lipophilicity (LogD) of candidate
compounds was also measured by the “shake-flask” method to estimate the brain permeability. As a
result, compound 8 was identified as the most promising irreversible MAGL inhibitor, and compounds
1
7 and 37 were identified as the most promising reversible inhibitors. Possible molecular interactions
between these lead compounds and MAGL, as well as the corresponding binding domain were also
1
1
18
11
identified through molecular docking studies. The corresponding C- and F-isotopologues 48 ([ C]8;
1
1
11
11
11
also known as [ C]MAGL-2-11), 49 ([ C]17; also known as [ C]PAD) and 50 ([ C]37; also known
1
8
as [ F]MAGL-4-11) were synthesized in excellent radiochemical yield, high molar activity and high
11
radiochemical purity, and with diverse radiolabeling strategies, including [ C]COCl cyclization, in-
2
1
1
11
18
loop [ C]CO fixation, C-methylation and F-fluorination. The pharmacokinetic profiles of these
2
radioligands including brain uptake, clearance and binding specificity were further investigated by PET
imaging and ex vivo whole-body distribution studies in rodents. The irreversible tracer 48 demonstrated
excellent BBB penetration and heterogenous regional brain uptake and high binding specificity, which
was in line with MAGL distribution. The reversible tracers 49 and 50 showed limited brain uptake, but
exhibited excellent binding specificity towards MAGL in the periphery as evidenced by ex vivo
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
blocking bio-distribution studies. We also performed PET experiments in Pgp/Bcrp knockout mice, in
which 49 and 50 exhibited substantially increased activity accumulation in the brain, indicating that the
underlying cause of their limited brain uptake was probably attributed to the significant interaction with
ABC efflux transporters. Therefore, further optimization is necessary to increase brain permeability by
minimizing the interaction with Pgp/Bcrp efflux transporters, and further validation in comprehensive
in vivo MAGL inhibitory experiments is also desirable to evaluate the therapeutic potential of these
novel molecules. Multiparameter optimization (MPO), namely the approach to simultaneously optimize
and balance basicity (pKa), lipophilicity (clogP and clogD), number of hydrogen bond donors (HBD),
polar surface area (tPSA) and molecular weight (MW), has demonstrated its usefulness in the
improvement of molecular properties and ADME profile, including brain penetration.42, 63-66 We will
utilize this strategy to design next-generation brain penetrant radioligands for reversible binding
mechanism. In conclusion, this work not only provides the first synergetic medicinal chemistry
approach to both irreversible and reversible MAGL inhibitors based on a unique piperazinyl azetidine
scaffold, but also paves the way towards the development of next generation reversible PET tracers for
translational PET imaging studies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EXPERIMENTAL SECTION
General consideration. The general procedure for experimental section was described previously41
with minor modification in this work. All the chemicals employed in the synthesis were purchased from
commercial vendors and used without further purification. Thin-layer chromatography (TLC) was
60
conducted with 0.25 mm silica gel plates ( F ) and visualized by exposure to UV light (254 nm) or
2
54
stained with potassium permanganate. Flash column chromatography was performed using silica gel
ACS Paragon Plus Environment

Page 29 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(particle size 0.040-0.063 mm). Nuclear magnetic resonance (NMR) spectra were obtained either on a
Bruker spectrometer 300 MHz. Chemical shifts (δ) are reported in ppm, and coupling constants are
reported in Hertz. The multiplicities are abbreviated as follows: s = singlet, d = doublet, t = triplet, q =
quartet, quint = quintet, sext = sextet, sept = setpet, m = multiplet, br = broad signal, dd = doublet of
doublets. For mass spectrometer measurement, the ionization method is ESI using Agilent 6430 Triple
Quad LC/MS. Lipophilicity was calculated by ChemBioDraw Ultra 14.0 (CambridgeSoft Corporation,
PerkinElmer, USA). All tested MAGL inhibitors showed high purity (≥95%) as determined by a
reverse-phase HPLC (Gemini ® 5 μm, NX-C18 110 Å column (3 mm ID × 150 mm). The lead
compounds 8, 17 and 37 did not show any promiscuous moieties in the Pan Assay Interference
Compounds Assay (PAINS) using two different in silico filters (http://www.swissadme.ch/index.php
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
7
and http://zinc15.docking.org/patterns/home/). Molar activity determinations are reported at the end
of synthesis, unless otherwise stated. The animal experiments were approved by the Institutional
Animal Care and Use Committee of Massachusetts General Hospital or the Animal Ethics Committee
at the National Institutes for Quantum and Radiological Science and Technology, National Institute of
Radiological Sciences. CD-1 mice (female; 7 weeks, 22-24 g), DdY mice (male; 7 weeks, 34-36 g),
Sprague-Dawley rats (male; 7 weeks; 210-230 g) were kept on a 12 h light/12 h dark cycle and were
allowed food and water ad libitum. Male Pgp/Bcrp knockout (Mdr1a/1b(-/-) and Abcg(-/-)) miceand
male wild-type mice (FVB) were purchased from Taconic (Hudson, NY).
Chemistry.
General procedure for synthesis of 1,1,1,3,3,3-hexafluoropropan-2-yl 3-(4-(thiazole-2-
carbonyl)piperazin-1-yl)azetidine-1-carboxylate (8). Under nitrogen atmosphere, a solution of 4-
ACS Paragon Plus Environment