Benzisothiazolinone Derivatives as Potent Allosteric Monoacylglycerol Lipase Inhibitors That Functionally Mimic Sulfenylation of Regulatory Cysteines

Article abstract of DOI:10.1021/acs.jmedchem.9b01679

We describe a set of benzisothiazolinone (BTZ) derivatives that are potent inhibitors of monoacylglycerol lipase (MGL), the primary degrading enzyme for the endocannabinoid 2-arachidonoyl-sn-glycerol (2-AG). Structure-activity relationship studies evaluated various substitutions on the nitrogen atom and the benzene ring of the BTZ nucleus. Optimized derivatives with nanomolar potency allowed us to investigate the mechanism of MGL inhibition. Site-directed mutagenesis and mass spectrometry experiments showed that BTZs interact in a covalent reversible manner with regulatory cysteines, Cys201 and Cys208, causing a reversible sulfenylation known to modulate MGL activity. Metadynamics simulations revealed that BTZ adducts favor a closed conformation of MGL that occludes substrate recruitment. The BTZ derivative 13 protected neuronal cells from oxidative stimuli and increased 2-AG levels in the mouse brain. The results identify Cys201 and Cys208 as key regulators of MGL function and point to the BTZ scaffold as a useful starting point for the discovery of allosteric MGL inhibitors.

Full text of DOI:10.1021/acs.jmedchem.9b01679

Subscriber access provided by - Access paid by the | UCSF Library
Article
Benzisothiazolinone derivatives as potent allosteric
monoacylglycerol lipase inhibitors that functionally
mimic sulfenylation of regulatory cysteines
Riccardo Castelli, Laura Scalvini, Federica Vacondio, Alessio Lodola, Mattia Anselmi,
Stefano Vezzosi, Caterina Carmi, Michele Bassi, Francesca Ferlenghi, Silvia Rivara, Ingvar
R. Möller, Kasper D. Rand, Jennifer Daglian, Don Wei, Emmanuel Yaw Dotsey, Faizy
Ahmed, Kwang-Mook Jung, Nephi Stella, Simar Singh, Marco Mor, and Daniele Piomelli
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01679 • Publication Date (Web): 12 Nov 2019
Downloaded from pubs.acs.org on November 12, 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 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Benzisothiazolinone derivatives as potent allosteric monoacylglycerol lipase inhibitors that
functionally mimic sulfenylation of regulatory cysteines
1
1
1,2
1
1
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
Riccardo Castelli, Laura Scalvini, Federica Vacondio, Alessio Lodola, Mattia Anselmi,
Stefano Vezzosi, Caterina Carmi, Michele Bassi, Francesca Ferlenghi, Silvia Rivara,1,2,* Ingvar
1
1
1
1,2
3
3
4
4
4
4
R. Møller, Kasper D. Rand, Jennifer Daglian, Don Wei, Emmanuel Y. Dotsey, Faizy Ahmed,
Kwang-Mook Jung, Nephi Stella, Simar Singh, Marco Mor, Daniele Piomelli4,6,7,8
4
5
5
1,2
1
Dipartimento di Scienze degli Alimenti e del Farmaco, Università degli Studi di Parma, Parco
Area delle Scienze 27/A, I-43124 Parma, Italy
2
Centro Interdipartimentale Biopharmanet-tec, Università degli Studi di Parma, Parco Area delle
Scienze, Tecnopolo Padiglione 33, I-43124 Parma, Italy
3
Department of Pharmacy, Universitetsparken 2, DK-2100 Copenhagen
4
Department of Anatomy and Neurobiology, School of Medicine, University of California, Irvine,
California 92697, United States
5
Department of Pharmacology, Psychiatry and Behavioral Sciences, University of Washington,
Seattle, WA 98195-7280, United States.
6
Departments of Biological Chemistry and Pharmacology, School of Medicine, University of
California, Irvine, California 92697, United States
7
Center for the Study of Cannabis, University of California, Irvine, California 92697, United States
8
Visiting Professor, Università degli Studi di Parma under the TeachInParma project.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 64
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
Abstract
We describe a set of benzisothiazolinone (BTZ) derivatives that are potent inhibitors of
monoacylglycerol lipase (MGL), the primary degrading enzyme for the endocannabinoid 2-
arachidonoyl-sn-glycerol (2-AG). Structure-activity relationship studies evaluated various
substitutions on the nitrogen atom and the benzene ring of the BTZ nucleus. Optimized derivatives
with nanomolar potency allowed to investigate the mechanism of MGL inhibition. Site-directed
mutagenesis and mass spectrometry experiments showed that BTZs interact in a covalent reversible
manner with regulatory cysteines, Cys201 and Cys208, causing a reversible sulfenylation known to
modulate MGL activity. Metadynamics simulations revealed that BTZ adducts favor a closed
conformation of MGL that occludes substrate recruitment. The BTZ derivative 13 protected
neuronal cells from oxidative stimuli and increased 2-AG levels in mouse brain. The results identify
Cys201 and Cys208 as key regulators of MGL function, and point to the BTZ scaffold as a useful
starting point for the discovery of allosteric MGL inhibitors.
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
Introduction
Monoacylglycerol lipase (MGL) is a serine hydrolase that cleaves medium- and long-chain
monoacylglycerols into fatty acid and glycerol, with the highest rate for arachidonic acid
1
derivatives. MGL expression is widespread in the body, and particularly high in the central
2
nervous system (CNS) where it is exclusively localized to presynaptic nerve terminals. In
peripheral tissues, MGL serves two main functions: it mediates the last step in hormone-dependent
3
lipolysis and contributes to the release of energy-rich fatty acids from triacylglycerol stores and,
secondly, it controls platelet aggregation by regulating the availability of non-esterified
4
arachidonate for thromboxane A biosynthesis. By contrast, the primary role of MGL in the CNS is
2
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
to catalyze the hydrolytic deactivation of 2-arachidonoyl-sn-glycerol (2-AG),5,6,7 an endogenous
agonist at CB and CB cannabinoid receptors and the most abundant endocannabinoid found in the
1
2
mammalian brain.8,9 2-AG is released upon demand from postsynaptic neurons and acts as a
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
retrograde messenger to modulate presynaptic activity by engaging CB cannabinoid receptors on
1
1
0
axon terminals. This signaling modality is thought to underpin most of the effects elicited by 2-
AG in neural tissue, including antinociception,11,12 neuroprotection,13,14 and reduced anxiety. In
addition to these central CB -mediated actions, 2-AG also acts on peripheral CB and CB receptors
15
1
1
2
to mediate, among other effects, immune modulation16,17 and inhibition of cancer cell growth.18,19
The first MGL inhibitor to be reported was the carbamate derivative URB602 1 (Figure 1), which
acts through a partially reversible, noncompetitive mechanism.20,21,22 More potent and selective
compounds were subsequently discovered, which differ mechanistically from URB602 in that they
inhibit MGL activity by forming covalent adducts with the enzyme’s catalytic serine (Ser122).
2
3
24
These molecules include carbamate derivatives such as JZL184 (2), PF-06795071 (3) and ABX-
2
5
1
431. Recently, reversible MGL inhibitors have also been described, such as benzoylpiperidine
2
6
27
derivatives and piperazinyl pyrrolidin-2-ones.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
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
O
O
O
CF3
OH
H
N
N
O
O N
O
2
N
O
H
N
N
H
OH
O
O
O
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
JZL184, 2
PF-06795071, 3
URB602, 1
F
O
S
N
O
N
S
S
N
O
N
S
S
disulfiram, 4
NAM, 5
octhilinone, 6
COOCH3
O
O
N
S
HO
7
pristimerin, 8
Figure 1. Chemical structures of current MGL inhibitors.
2
8
MGL is a member of the / hydrolases superfamily of enzymes and requires a Ser/His/Asp triad
for its catalytic activity.29,30 An interesting structural feature of this enzyme is the presence of a lid
domain that surrounds and hides its catalytic core. Within this domain, a region comprising helix
4 and the loop connecting helices 4 and 5 forms a flexible, hydrophobic U-shaped domain,
which may serve both as a portal restricting or allowing the access of substrate into the catalytic
site, and as a lipid-binding surface that reversibly interacts with membranes, where the substrate
resides (Figure 2). Another characteristic of MGL is the presence of three cysteine residues –
Cys242, Cys201 and Cys208 – that modulate catalytic activity in important ways. Cys242 is located
close to the catalytic Ser122 and is directly involved in catalysis, as both its mutation and its
3
1
covalent modification strongly reduce MGL’s hydrolytic activity. Indeed, computational and mass
spectrometry studies have shown that the cysteine-reactive agent N-arachidonylmaleimide (NAM,
5
, Figure 1) forms a covalent adduct with this cysteine residue. 32,33,34 It is likely that other Michael
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acceptors and disulfide-containing compounds, such as tetraethyluram disulphide (disulfiram, 4,
3
5
Figure 1), which inhibit MGL activity, act through a similar mechanism. Unlike Cys242, Cys201
and Cys208 lie at a distance from the active site. Cys201 is located on the loop connecting helices
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 and 6, while Cys208 is on the first turn of helix 6, close to the putative position of the
membrane bilayer (Figure 2). These cysteine residues are known allosteric modulators of rat and
human MGL activity.36 When rat MGL is exposed to relatively low concentrations of the
intracellular messenger hydrogen peroxide,37,38 Cys201 and Cys208 undergo reversible
sulphenylation and the enzyme experiences a parallel decrease in catalytic activity.39 This
modulatory effect may be of physiological significance and it has been linked to the ability of
covalent modifications of Cys201 and Cys208 to affect the interfacial activation of MGL by
promoting the closure of the lid domain and reducing substrate recruitment.40 The relevance of
these regulatory cysteines has been further highlighted by the discovery that several reversible
MGL inhibitors require Cys201 and Cys208 for activity. These include octyl-isothiazolinone
octhilinone, 6) and octyl-benzisothiazolinone (7),41,42 as well as the sesquiterpene pristimerin (8).36
(
Figure 2. (A) The crystal structure of human MGL with the lid domain in its open conformation
(PDB 3HJU) is shown in grey with the U-shaped access funnel to the catalytic site in orange. The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
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
membrane profile is represented with light-pink spheres. (B) Helix α4 and the loop connecting
helices α4 and α5 delineate a wide U-shaped structure that delimits the substrate access to the
catalytic site.
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 findings outlined above suggest that molecular probes that interact with regulatory cysteines
Cys201 and Cys208 may offer important insights on the functioning of MGL as well as an
alternative strategy to inhibit its activity. To test this idea, in the present study we introduced
selected modifications in the structure of the BTZ derivative 7 with three interrelated objectives: i)
enhance recognition by MGL and improve inhibitory potency; ii) define the molecular mechanism
through which BTZs interact with MGL; iii) identify new probes that may guide the discovery of
reversible allosteric MGL-targeting agents with potential therapeutic utility.
Chemistry
43
The synthesis of BTZs was performed as outlined in Schemes 1-3. Compounds 10, and 10a were
obtained via alkylating commercially available 9 with potassium carbonate and the appropriate
iodoalkane. Compound 11 was obtained from 10a by reacting the 6-chlorohexyl moiety with
piperidine (Scheme 1). Compounds 13-16 were synthesized from freshly prepared 2-
chlorosulfenylbenzoyl chloride 12,44 with the respective amine. 2-Phenylethyl isothiazolinone 18
was prepared according to a previously reported procedure.45
Scheme 1a
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
a
NH
N R
S
S
9
10: R= methyl
0a: R= 6-chlorohexyl
11: R= 6-(piperidin-1-yl)hexyl
1
b
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
O
O
c
Cl
N R
S
S
Cl
13: R= phenethyl
14: R= 2-(naphthalen-1-yl)ethyl
15: R= 2-(naphthalen-2-yl)ethyl
1
2
1
6: R= cyclohexyl
O
N
O
H
d
S
S
N
H
N
S
18
O
17
a
Reagents and conditions: a) K CO , alkyl halide, MeCN, rt, 16 h; b) piperidine, K CO , KI,
2
3
2
3
MeCN, 60 °C, 16 h; c) amine, anhydrous TEA, DCM rt, 16 h; d) SO Cl , anhydrous DCE 0 °C, 2 h.
2
2
Phenyl-substituted BTZs 24-30 were synthesized with a unified approach as outlined in Scheme 2.
Scheme 2a
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 64
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
O
O
O
O
O
OMe
N
OMe
N
a
b
c
OMe
OH
OH
SR'
S
R
R
R
R
S
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
22a: R= 5-nitro; R'=H
1
9a: R= 5-nitro
9b: R= 4-nitro
19c: R= 5-methoxy
9d: R= 4-methoxy
9e: R= 3-methyl
20a: R= 5-nitro
20b: R= 4-nitro
20c: R= 5-methoxy
20d: R= 4-methoxy
20e: R= 3-methyl
21a: R= 5-nitro
21b: R= 4-nitro
21c: R= 5-methoxy
21d: R= 4-methoxy
21e: R= 3-methyl
22b: R= 4-nitro; R'=H
22c: R= 5-methoxy; R'=H
22d: R= 4-methoxy; R'=H
1
1
1
2
2e: R= 3-methyl; R'=SCONMe2
O
O
d
Cl
e
N
2
2a-e
S
SR'
R
R
24: R= 5-nitro
2
2
3a: R= 5-nitro; R'=H
3b: R= 4-nitro; R'=H
23c: R= 5-methoxy; R'=H
23d: R= 4-methoxy; R'=H
23e: R= 3-methyl; R'=SCONMe2
2
2
2
2
5: R= 6-nitro
6: R= 5-methoxy
7: R= 6-methoxy
8: R= 7-methyl
f
f
29: R= 5-hydroxy
30: R= 6-hydroxy
aReagents and conditions: a) base, N,N-dithiocarbamoylchloride, anhydrous DMF; b)
diphenylether; W irradiation c) NaOH, MeOH/H O 2:1 v/v, reflux; d) SOCl , reflux, 6 h; e) i) 2-
2
2
phenylethylamine, anhydrous TEA, THF ii) I , anhydrous THF, 0 °C to rt, 20 h; f) BBr , anhydrous
2
3
benzene, 0 °C to reflux, 3 h.
Commercially available methyl salicylates 19a-e were reacted with N,N-dithiocarbamoyl chloride
in the presence of a base, either DABCO (19a and 19b) or NaH (19c-e) in DMF. The generated O-
thiocarbamates 20a-e were subjected to an intramolecular aryl-group migration (Newman-Kwart
4
6
rearrangement) induced by microwave irradiation at high temperature. The methyl ester group and
the S-thiocarbamate groups of 21a-d were hydrolyzed contextually by exposure to NaOH, leading
to formation of 2-mercapto benzoic acids 22a-d. In the case of compound 21e, hydrolysis of the S-
thiocarbamoyl (-SCONMe ) group failed under different reaction conditions; it was therefore
2
carried along in the synthesis. Substituted benzoic acids 22a-e were then treated with SOCl and the
2
crude, unpurified acyl chlorides 23a-e thus obtained were reacted with 2-phenylethylamine,
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
followed by oxidative cyclization with iodine in a one-pot fashion, to give final compounds 24-28.47
Compounds 26 and 27 were further elaborated by removing the methoxy group by exposure to
boron tribromide in benzene, to yield final products 29 and 30, respectively.48
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
Synthesis of BTZs bearing a benzylamino side chain is summarized in Scheme 3.
Scheme 3a
O
O
O
R
R'
N N
Boc
N NH
N N
n
R
a
a
S
N
O
S
R
S
N
31
S
38a: R= Boc; n=5
b
3
3
4
8b: R= Boc; n=6
9: R= H; n=5
0: R= H; n=6
3
2a: R= H; R'= Boc
b
32b: R= COOMe; R'= Boc
32c: R= COOt-Bu; R'= Boc
32d: R= Ph; R'= Boc
b
c
3
3: R= H; R'= H
37: R= H; R'= Me
34: R= COOMe; R'= H
5: R= COOH; R'=H
6: R= Ph; R'= H
3
3
aReagents and conditions: a) KHMDS, alkyl halide, anhydrous DMF, 0 °C to rt, 16 h; b) TFA,
DCM, 0 °C to rt, 4 h; c) MeI, imidazole, rt, 16 h.
4
9
BTZ 31 was treated with KHMDS and the proper alkyl halide to generate intermediates 32a-d,
from which, upon exposure to an equivolume mixture of trifluoroacetic acid and dichloromethane,
the tert-butoxy carbonyl protecting group was removed to furnish products 33-36. Compound 37
was obtained by methylation of 33 with methyl iodide and imidazole. Symmetrically substituted 39
and 40 were obtained from 31 employing the same sequence of alkylation with either 1,5-
diiodopentane (38a) or 1,6-diiodohexane (38b), followed by TFA-induced Boc-deprotection, to
yield 39 and 40, respectively.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 64
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
Results and discussion
Starting from the potent BTZ derivative 7,41 we evaluated different side chains on the nitrogen
atom, as well as additional substituents on the BTZ nucleus. The new derivatives were tested for
their ability to inhibit recombinant rat MGL overexpressed in E. coli. We first focused on the effect
exerted by substituents on the nitrogen atom characterized by different size, shape and lipophilicity.
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
Half-maximal inhibitory concentration (IC ) values for BTZs belonging to this first set are reported
50
in Table 1 along with the IC of reference compound 7.
5
0
Table 1. Inhibition of Purified Recombinant Rat MGL by N-Substituted Benzisothiazolinones.
O
O
O
N R
N
R
N
S
S
S
7
, 9-11, 13-16, 33-37
18
24-30
a
Compd.
R
IC (nM)
50
7
9
n-octyl
H
59b
306.5 ± 32.7
59.5 ± 5.8
1
1
0
1
Me
195.2 ± 12.3
34.1 ± 2.0
1
3
1
1
4
5
21.8 ± 5.2
54.2 ± 15.7
ACS Paragon Plus Environment

Page 11 of 64
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
1
6
55.5 ± 7.4
1
2
8
4
-
758.6 ± 49.8
163.4 ± 2.8
5-NO2
6- NO2
5-OMe
6-OMe
7-Me
5-OH
6-OH
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
25
237.9 ± 10.3
568.0 ± 55.3
512.2 ± 64.4
13069.5 ± 523.9
135.0 ± 3.8
2
2
6
7
28
2
3
9
0
2547.0 ± 879.7
3
3
3
4
19.5 ± 0.8
93.9 ± 5.5
3
3
5
6
277.7 ± 1.9
322.1 ± 18.7
189.4 ± 39.7
37
aIC50 values were calculated from dose-response inhibition curves of 2-OG hydrolysis, on cell
homogenates of HeLa cells expressing recombinant rat MGL, and 10 min pre-incubation with test
compound.22,71 Results are expressed as mean ± SEM (n = 3-4). Ref.41.
b
The flexible octyl chain of BTZ 7 was initially replaced with more rigid substituents, as in
compounds 14, 15 and 16 in which the naphthyl-ethyl chains and the cyclohexyl ring appear to be
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
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
well tolerated. Size reduction, obtained with a methyl (10) or a phenylethyl (13) group on the BTZ
nitrogen atom, allowed to maintain the same level of inhibitory potency, suggesting that the enzyme
can tolerate broad changes in steric bulk and lipophilicity of the N-side chain. By contrast, more
hydrophilic compounds, such as the piperidino-hexyl derivative 11 or the unsubstituted BTZ 9,
were characterized by a significant reduction of their inhibitory potency. The absence of a
substituent on the nitrogen atom in compound 9 allows a prototropic equilibrium which might
reduce the availability of the reactive amide form of the sulfenamide warhead. The isothiazolinone
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
8 carrying a phenethyl chain was significantly less potent than its BTZ analog 13.
We selected the N-phenethyl-BTZ scaffold of 13 to investigate the effect of substituents on the
benzene portion of the bicyclic nucleus. Electron-withdrawing or donating groups were introduced
in positions conjugated with the sulfur atom or the carbonyl group, to modulate the reactivity of the
sulfenamide warhead. We also evaluated the influence of steric hindrance close to the reactive
sulfur atom (compounds 24-30 in Table 1). Insertion of substituents on the benzene ring led to a
general loss of activity, with no clear influence of the electronic effect on inhibitory potency. In
fact, the most potent derivatives 24 and 29 carry an electron-withdrawing and donating substituent,
respectively, in position 5. While the nitro group in position 6 (25) maintained some potency, the 6-
hydroxy substituent (30) led to one of the highest drops of activity in this series. Compound 28 was
the least potent, highlighting the negative effect exerted by a substituent in position 7, close to the
reactive sulfur of the BTZ nucleus. The steric hindrance of the methyl group likely affects the
reactivity of BTZ, reducing the rate of nucleophilic substitution by the thiol group of targeted
cysteines. The results highlight that structural modulation on the benzene ring influences activity in
an unpredictable manner. At the same time, the loss of potency caused by introduction of steric
hindrance confirms the important role played by the sulfur atom.
To inform subsequent SAR studies on compound 13, we examined its interaction with MGL using
two complementary approaches: site-directed mutagenesis and LC/MS analysis. First, we generated
HeLa cells expressing wild-type rat MGL or MGL mutated on Cys201, Cys208 or Cys201 plus
ACS Paragon Plus Environment

Page 13 of 64
Journal of Medicinal Chemistry
1
2
3
Cys208. Cell extracts were evaluated for MGL activity using a standard enzyme assay (see Methods
for details). The results showed that compound 13 was significantly more potent at inhibiting the
wild-type form of MGL than it was at inhibiting mutated forms of the enzyme, with IC values
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
5
0
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
measured for C201A, C208A and C201/208A mutants of 242.9, 72.6, and 604.7 nM, respectively
4
1
(
IC = 34.1 nM for wild-type MGL) (Figure 3A). As expected from our previous studies, MGL
50
inhibition by compound 13 was fully reversed upon rapid dilution. By contrast, the non-selective
serine-targeting inhibitor MAFP (methyl arachidonyl fluorophosphonate) blocked MGL activity in
an irreversible manner (Figure 3B).
A
B
O
NH
S
S
MGL
C
Figure 3. (A) Effect of cysteine mutations on rat MGL inhibition by compound 13. Results are
expressed in percentage of control activity (DMSO, final concentration 1%) (mean ± SEM, n = 3-
5
). Wt = wild type. (B) Rapid dilution assays of purified rat MGL in the presence of vehicle (Veh,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
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
DMSO, final concentration 2%), compound 13 (10 μM final concentration), or MAFP (23 nM final
concentration). Results are expressed as MGL activity after the indicated incubation time. (C) Top:
deconvoluted ESI mass spectrum of wild-type human MGL (average mass: 34084  2 Da, n=3).
Bottom: deconvoluted ESI mass spectrum of wild-type MGL incubated in the presence of a four-
fold molar excess of BTZ 13. Mass shifts of 254 Da corresponds to the conjugation of one (average
mass: 34338  2 Da, n=3) or two (average mass: 34592  2 Da, n=3) molecules of BTZ.
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
Next, we incubated recombinant wild-type human MGL (Figure S1) in the presence of BTZ 13
(MGL:13 1:4 molar ratio, 45 min). The results show that, after exposure to BTZ, the average mass
of intact apo-hMGL (34084  2 Da, n=3) was shifted to a value of 34338  2 Da (n=3), compatible
with the formation of a 1:1 complex with BTZ 13 (average mass: 255.335) (Figures 3C, S2 and S3).
A second signal was also observed at a higher m/z (mass-to-charge ratio) (34592  2 Da), which
could be attributed to the conjugation of two molecules of BTZ to MGL, but its intensity was ~15-
20 % that of the 1:1 MGL-13 adduct. Then, apo-MGL and MGL incubated with 13 were digested
with pepsin to identify specific conjugation site(s). Following apo-MGL incubation, peptides
containing three out of the four cysteine residues of human MGL (Cys32, 201 and 242) could be
identified (Figure S4A). The short amino acid sequence KVCF, which includes the regulatory
Cys208 of MGL, was not covered by peptic peptides despite an overall sequence coverage of
9
8.7%. In MGL incubations with BTZ 13, Cys32- and Cys242-containing peptides were identified,
as in apo-MGL, but identification of Cys201-containing peptides was not possible indicating
modification of this peptide (Figure S4B). In conjunction, a new peptide was identified with an
average mass and MS/MS fragmentation pattern compatible with that of the Cys201-containing
peptide (aa 193-205 in native human MGL) covalently modified by BTZ (Table S1 and Figure S5).
The formation of a reducible disulfide bond between BTZ 13 and the cysteine residue was
supported by the fact that application of a reducing agent resulted in the disappearance of the BTZ-
ACS Paragon Plus Environment

Page 15 of 64
Journal of Medicinal Chemistry
1
2
3
modified peptide, but not of the Cys201-containing peptide in its reduced form (Table S1). These
results confirm the mechanism of inhibition of MGL by BTZ 13 via interaction with Cys201, and
support a covalent modification of this residue.
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
To investigate the reactivity of BTZs toward biologically relevant thiols and their ability to promote
the chemical transformation of their counterpart, compound 13 was incubated with two equivalents
of glutathione (GSH) in physiological conditions (PBS buffer, pH 7.4, 37 °C). The transformations
undergone by BTZ 13 are depicted in Figure 4, while the extracted ion chromatograms from the
HR-MS full scan analysis are reported in Figure S6. According to reaction A, BTZ 13 reacted with
GSH to form the conjugated derivative 41, mimicking the reaction occurring with cysteines from
MGL. Mixed disulfide 41, upon reaction with another GSH molecule, generated the substituted 2-
mercaptobenzamide 42 and oxidized glutathione (GSSG). BTZ 13 could also react (reaction B in
Figure 4) with the 2-mercaptobenzamide 42 to generate the symmetric disulfide 43.
A
O
O
O
N
H
GSH
GSH
+
GSSG
N
N
H
S
S
S
SH
42
G
41
1
3
O
B
N
O
O
H
S
S
+
N
N
H
S
H
N
SH
13
42
O
43
Figure 4. Reactions occurring between BTZ 13 and glutathione (GSH) in solution when present in
a 1:2 molar ratio.
Table 2 reports the concentrations of unreacted GSH, oxidized dimeric GSH derivative GSSG, and
residual BTZ 13, and the peak area corresponding to the conjugate between BZT and GSH (41,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
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
Figure 4) obtained from a solution of GSH 200 M and BTZ derivative 13 100 M after 60 min of
reaction time. BTZ 13 fully reacted with GSH generating the covalent adduct 41 and producing a
significant amount of oxidized dimeric GSSG.
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
Table 2. Reactivity of BTZ 13 with Glutathione.a
Compd.
GSH
M)b
GSSG
% oxid. c BTZ
13 Conjugate 41d
(
(M)b
(M)
1
3
73.4 ± 4.7 59.3 ± 1.1 61.8
< LODe
721
a
b
Data collected at t = 60 min. Starting concentrations: GSH = 200 M; BTZ 13 = 100 M. Results
c
are expressed as mean ± SEM (n = 3-4). % of oxidized GSH, calculated as (2[GSSG]/[GSH] +
d
2[GSSG])  100. Value corresponding to peak area of UV chromatogram; a quantification was not
e
possible due to lack of standards. Below limit of detection (LOD).
Supporting the possibility that the interaction of compound 13 with MGL mimics at the functional
level hydrogen peroxide-induced sulfenylation, metadynamics simulations revealed that the
formation of an adduct between 13 and MGL perturbs the conformational equilibrium between
4
0
open and closed states of the lid domain, as previously observed for hydrogen peroxide. In fact,
both human and rat enzymes are characterized by a dampened activity under oxidative
conditions39,40 and free energy studies provided a plausible description of the mechanism laying
behind the oxidative-dependent MGL modulation. Metadynamics simulations suggested that
Cys201, reversibly converted into sulfenic acid by oxidative stimuli (e.g., hydrogen peroxide),
hampers the recruitment of the substrate from the membrane by affecting the interfacial activation
of MGL.40 To test if BTZ derivatives might modulate MGL through a similar mechanism, we
performed well-tempered (wt) metadynamics simulations of human MGL modeled with Cys201
covalently modified by compound 13, in the presence of a membrane model. The Collective
Variables (CV1 and CV2) selected to drive the transition between the open and closed states of the
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
lid domain were distances between the centers of mass of Cα of residues lining the rims of the lid
domain (see Methods for details). This region, representing a membrane-related hydrophobic gate
that shapes the active site access channel of the substrate, is stabilized in an open conformation in
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
4
0
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
presence of a membrane model, allowing the recruitment of 2-AG from the membrane. The free
energy surface (FES) profile obtained for MGL covalently bound to BTZ 13 revealed only one
energy minimum corresponding to the closed state characterized by a reduced distance between the
lid domain rims, driven by the simultaneous reduction of CV1 and CV2 values (Figure 5). These
results suggest that, similarly to what had been observed for the sulfenylated model of the
enzyme,40 modification of Cys201 through covalent interaction with compound 13 hampers
catalytic activity by promoting the closure of the active site access channel.
Figure 5. Well-tempered metadynamics simulation of human MGL (PDB 3HJU) with Cys201
covalently bound to BTZ 13. (A) Free energy surface. A black diamond marks CV1 and CV2
values assumed by the system in its open conformation, measured in the crystal structure 3HJU and
corresponding to the starting structure of the metadynamics simulation. (B) Starting conformation
of wt metadynamics simulations. Compound 13 covalently bound to Cys201 is represented with
green carbons. The rims of the lid domain are depicted with purple cartoons. (C) A conformation
corresponding to the free energy minimum of the metadynamics simulation, showing a marked
decrease of both CV values and thus a reduced distance between the rims of the lid domain.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 64
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
Having established that 13 binds to Cys201, but not to active-site Cys242 or peripheral Cys32, we
carried out a series of covalent docking studies to generate novel hypotheses about the nature of the
interaction between 13 and MGL (Figure 6). A crystal structure of human MGL in which the lid
domain is in an open conformation (pdb 3HJU) was used for covalent docking. We modelled the
formation of a disulfide adduct between the BTZ derivative and the thiol group of Cys201, with
concomitant opening of the isothiazolinone ring and generation of a benzamide derivative. Analysis
of docking solutions suggests the existence of four potential strategies for BTZ recognition by MGL
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 6A): the formation of i) a hydrogen bond with the carbonyl group of Pro198, as occurring
for BTZ 13; ii) polar interactions with Arg87; iii) hydrophobic interactions with the lipophilic
residues surrounding Cys201; and iv) a possible bidentate interaction with both Cys201 and
Cys208.
Figure 6. Covalent docking of BTZ 13 (panel A) and 33 (panel B) on Cys201 of human MGL
(PDB 3HJU). The molecular surface of MGL in the proximity of the bound ligand is represented in
gray. Dashed lines indicate hydrogen bonds.
Following these indications, we synthesized 33-37 (Table 1). Replacement of the phenethyl chain
on the BTZ nitrogen with a benzylamino group (33) resulted in a small but significant improvement
in potency. Compound 33 forms a hydrogen bond between the NH group from the benzylamino
ACS Paragon Plus Environment

Page 19 of 64
Journal of Medicinal Chemistry
1
2
3
side chain and the backbone carbonyl of Pro198, undertaking an additional polar interaction with
Arg87 (Figure 6B). Consistently with a role in the recognition event by MGL, methylation of the
benzylamino nitrogen significantly reduced potency measured for compound 37. Attempts to
establish additional interactions with polar amino acids (e.g., Arg87) or with the hydrophobic
surroundings of Cys201 were performed by inserting suitable substituents on the para position of
the benzylamino side chain. However, these structural modifications did not improve potency:
indeed, inhibitory potency was maintained in derivative 34, but was progressively decreased in
compounds 35 and 36.
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
Docking simulations also highlighted the possibility to accommodate the covalent adducts within
the region spanning from Cys201 to Cys208. This observation, coupled with the results of
mutagenesis and LC/MS studies, which suggested the participation of both Cys201 and Cys208 to
the covalent interaction with BTZs, prompted us to design compounds able to interact with both
regulatory cysteines. Based on this premise, we performed an induced fit docking (IFD) study for
symmetric BTZ derivatives based on the scaffold of inhibitor 33, looking for the optimal linker
length between the two reactive portions. The results of the IFD study are represented in Figure 7
for compound 39. The flexible 5-methylene chain occupies the region between Cys201 and Cys208,
while the sulfur atoms of the BTZ rings are close to the thiol groups of the two cysteine residues.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
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
Figure 7. IFD solution obtained for compound 39 spanning the region from Cys201 to Cys208 of
human MGL (PDB 3HJU); orange dashes indicate the distance between the sulfur atoms of the
benzisothiazolinone rings and those of Cys201 and Cys208.
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
Compounds 39 and 40, with a 5- and 6-carbon polymethylene chain, respectively, showed
significant MGL inhibitory potencies, with IC50 values of approximately 20 nM (Table 3).
Mutagenesis data for the most potent BTZ derivatives highlight that inhibitory activity was
significantly decreased when Cys201 or Cys208 were mutated to alanine, supporting the
involvement of these residues in the inhibitory mechanism (Table 3). Inhibitory potency was further
reduced when both cysteines were mutated. This behavior is consistent with a cooperative
interaction between the two cysteines, and reproduces what has been previously observed after
treatment of MGL with hydrogen peroxide. In fact, hydrogen peroxide oxidized both Cys201 and
Cys208, leading to a reduction of MGL activity, impairment of 2-AG degradation and enhancement
3
9
of 2-AG-mediated signaling in brain neurons. IC values measured for compounds 39 and 40 in
5
0
the presence of the C201&208A mutation were twice as high as those of the monomeric derivatives
3 and 33, supporting a preferential interaction of symmetrical compounds with these two residues.
1
Table 3. Inhibition of Purified Recombinant Rat MGL with Mutation of Cys201 and Cys208.
a
IC (nM)
50
Compd. Structure
Wild type
C201A
C208A
C201&208A
1
3
34.1 ± 2.0
242.9 ± 18.0 72.6 ± 30.3
604.7 ± 10.2
3
3
3
9
19.5 ± 0.8
11.5 ± 1.5
54.4 ± 4.1
44.9 ± 8.7
782.0 ± 77.5
O
O
229.8 ± 69.0 219.8 ± 33.9 1460.7 ± 465.0
H
N
NH
(CH2)5
N
N
S
S
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
O
O
H
N
NH
(CH2)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
N N
4
0
16.8 ± 0.4
114.3 ± 8.9
82.7 ± 11.6
2079.6 ± 40.0
S
S
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
aResults are expressed as mean ± SEM (n = 3-4).
Next, we asked whether cysteine-targeting probes such as BTZs might inhibit MGL activity and
increase 2-AG levels in intact biological systems. First, we incubated Neuro-2a cells in cultures
with hydrogen peroxide, and assessed cellular damage by evaluating lactate dehydrogenase (LDH)
release into the culture medium (Figure 8A). For these experiments, we selected an extracellular
concentration of hydrogen peroxide (300 M) that was expected to yield intracellular hydrogen
39
peroxide levels of approximately 30 M, or three times the IC of hydrogen peroxide for MGL.
5
0
Hydrogen peroxide produced significant cellular toxicity, and this effect was heightened by CB1
cannabinoid receptor blockade with the selective inverse agonist rimonabant, but not by CB2
blockade with the selective antagonist AM630. This result is consistent with the protective action
5
0,51
exerted by CB receptor activation on oxidative damage.
LDH release was reduced by treatment
with the irreversible MGL inhibitor JZL184 (1 M) and the same effect was produced by BTZ 13
1 M). The reduction of hydrogen peroxide toxicity produced by compound 13 or JZL184 was
1
(
prevented by concomitant application of the CB antagonist rimonabant.
1
We then assessed the systemic availability of compound 13 following administration of a single 10
mg/kg dose (i.p.) in mice. Maximal plasma levels were reached 15 min after dosing and rapidly
declined thereafter (Figure 8B). A similar profile was observed in brain (Figure 8C), though access
to this organ was clearly limited (brain/plasma ratio ≈ 4.5%). In a parallel experiment, we
determined the ability of 13 (10 mg/kg, i.p.) to impair brain 2-AG degradation. LC/MS analyses
showed that 2-AG levels were increased in mice treated with the inhibitor, an effect that was
statistically detectable 2 h after administration (Figure 8D). The temporal discrepancy between 2-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
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
AG accumulation (maximal at 2 h) and brain levels of 13 (maximal at 15 min) may be accounted
for by slow accumulation of 2-AG in the face of persistent MGL inhibition by compound 13.
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
80
**
Plasma
60
4
0
0
0
2
0
15 30 45 60 75 90 105 120
Time (min)
A
B
4
3
2
1
0
*
Brain
80
60
*
4
2
0
0
0
0
15 30 45 60 75 90 105 120
Time (min)
Veh
Compound 13
C
D
Figure 8. (A) Effects of H O (300 M), alone or combined with CB antagonist rimonabant (Rim),
2
2
1
CB antagonist AM630, MGL inhibitor JZL184 (1 M) or compound 13 (1 M), on lactate
2
dehydrogenase (LDH) release in Neuro-2a cells. (B) and (C) Plasma and brain concentrations of
compound 13 after administration of 10 mg/kg to mice. *P<0.05 and **P<0.01 by two-tailed
Student’s t test or one-way ANOVA followed by Dunnett’s multiple comparison test. (D) Effect of
compound 13 (10 mg/kg) on 2-AG levels in mice brain 2 h after i.p. injection.
Finally, we investigated the interactions of compound 13 with a select series of potential off-targets.
We first tested whether the compound interacts (either as an agonist or an antagonist) with proteins
that feature free cysteines on their molecular surface, such as the transient receptor potential cation
ACS Paragon Plus Environment

Page 23 of 64
Journal of Medicinal Chemistry
1
2
3
channel subfamily A member 1 (TRPA1)52,53 and the transient receptor potential cation channel
subfamily V member 1 (vanilloid receptor, TRPV1).54 Table 4 reports % activation and
inactivation of human TRPA1 and TRPV1 channels produced by compound 13 at different
concentrations in an aequorin-based calcium mobilization assay. The compound showed little or no
agonist activity on either channel, and marginal antagonist activity on TRPA1 at concentrations
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
>
500 nM (25 times greater than those needed to block MGL).
Table 4. Activation and Inhibition of Human TRPA1 and TRPV1 Ion Channels Exerted by
Compound 13 and Binding Affinity of Compound 13 for rat cannabinoid receptors.a
Conc (nM)
hTRPA1 hTRPV1
rCB1
rCB2
Agonist assay
% activation
1
00
1000
000
Antagonist assay
-0.15
1.74
0.21
0.59
1.78
2
19.30
% inhibition
Binding assayb
% bound radioligand
5
5
0
-22.76
-9.82
-10.75
8.55
100.59
106.32
91.76
100.41
119.67
95.21
00
55.97
71.54
1000
a
b
Each concentration was tested in duplicate. Reported is the % bound radioligand in the presence
of compound 13 at the specified concentration.
To determine if compound 13 can affect the activity of other components of the endocannabinoid
system, we tested its affinity for rat CB and CB cannabinoid receptors, as well as the ability to
1
2
inhibit various enzymes involved in the metabolism of 2-AG or anandamide, i.e.: i. rat diglyceride
lipase- (DGL-), the serine hydrolase that catalyzes mGluR5-operated 2-AG formation in
spines;55,56 ii. mouse alpha/beta hydrolase domain containing 6 (ABHD6); and iii. rat fatty acid
ACS Paragon Plus Environment
57

Journal of Medicinal Chemistry
Page 24 of 64
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
58
amide hydrolase (FAAH). Compound 13 was devoid of significant binding at either rat CB or
1
CB receptors at the concentrations tested (50, 500 and 1000 nM) (Table 4). Figure 9 depicts the
2
inhibition curves of rat DGL- (left panel) and rat FAAH (middle panel) and the effect of
compound 13 on mouse ABHD6-catalyzed 2-AG hydrolysis (right panel). Compound 13 had no
effect on DGL- activity at any of the concentrations tested and only modestly inhibited ABHD6
and FAAH, with IC values of 3.58 and 9.38 M, respectively.
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
0
1
20
00
80
60
40
20
0
1
40
100
1
120
100
80
8
6
4
2
0
0
0
0
0
60
40
20
0
mABHD6
-
9
-8
-7
-6
-5
-4
-9
-8
-7
-6
-5
-4
-9
-8
-7
-6
-5
-4
Log [Compound 13]
Log [Compound 13]
Log [Compound 13]
Figure 9. Effects of compound 13 on diglyceride lipase (left) and FAAH (center) activities from rat
brain homogenate, and mouse ABHD6 activity from ABHD6-expressing HEK cell lysate (right).
Results are expressed as mean ± SEM; n = 3-6, with each assay was performed in duplicate.
Conclusions
We describe a series of BTZ derivatives that act as allosteric inhibitors of MGL activity by
covalently and reversibly interacting with Cys201 and Cys208, two regulatory cysteines located in
close proximity of the lid domain of this serine hydrolase. Metadynamics simulations revealed that
BTZ-modified MGL tends to assume a closed conformation that hampers substrate recruitment into
the enzyme active site. Focused SAR investigations led to the potent derivative 13, which showed
neuroprotective activity in vitro and was able to elevate brain 2-AG levels in vivo. Pharmacokinetic
profiling in mice shows a rapid access of the compound to plasma and brain after i.p.
ACS Paragon Plus Environment

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
administration, though further work is needed to improve plasma life-time and brain access for this
class of agents.
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
BTZ derivatives act as general cysteine-targeting agents and have been shown to covalently modify
multiple enzymes.59,60,61 Nevertheless, we found that compound 13 does not significantly interfere
with the activity of two cysteine-rich proteins (TRPAV1 and TRPA1) and of several key
components of the endocannabinoid system (DGL-, ABHD-6, FAAH). Four additional lines of
evidence indicate that 13 interacts selectively with MGL: i) its high inhibitory potency; ii)
mutagenesis experiments highlighting the involvement of Cys201 and Cys208 in its inhibitory
mechanism; iii) mass spectrometry data indicating interaction of 13 with Cys201, but not other
cysteines (Cys32 and Cys242); iv) SAR explorations clarifying the importance of lipophilic
substituents for productive interactions with the environment surrounding the modulatory cysteine
residues.
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
In sum, our results show that it is possible to develop small-molecule allosteric MGL inhibitors that
recapitulate the ability of the intracellular second messenger, hydrogen peroxide, to modulate MGL
activity and enhance 2-AG mediated signaling. It is tempting to speculate that a similar strategy
might be deployed to modify the activity of other proteins that contain modulatory cysteines, such
as peroxiredoxins and NRF2.
Experimental section
Chemistry
All chemicals were used as received unless stated otherwise. All reactions were performed under a
steady overpressure of nitrogen delivered through a balloon. Tetrahydrofuran was distilled over
sodium/benzophenone prior to use. Anhydrous solvents such as dichloromethane, N,N-
dimethylformamide, toluene and acetonitrile were directly purchased from an appropriate vendor,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
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
and were supplied stored over 3 Å molecular sieves and packed under Ar. They were subsequently
manipulated by syringe under a steady pressure of nitrogen. Column chromatography was
performed on silica gel 60 (0.040-0.063 mm) under forced flow of the appropriate solvent mixtures.
TLC analysis was conducted on HPTLC aluminum sheets (Sigma-Aldrich, silica gel 60, F ),
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
2
54
compounds were visualized by UV absorption (245 nm) and/or by dipping in a solution of
NH ) Mo O ·4 H O 25 g/L and (NH ) Ce(SO ) ·2 H O 10 g/L, in 1 L of 10% aqueous H SO .
(
4 6
7
24
2
4 4
4 4
2
2
4
Microwave reactions were conducted using a CEM Discover synthesis unit (CEM Corp., Matthews,
NC). Melting points were not corrected and were determined with a Gallenkamp melting point
1
apparatus. H-NMR spectra were recorded with a Bruker AV 300 or 400 MHz. Chemical shifts (δ
1
scale) are reported in parts per million (ppm), referenced to TMS or the residual solvent signal. H-
NMR spectra are reported in the following order: multiplicity, number of protons and approximate
coupling constant (J value) in hertz (Hz); signals were characterized as s (singlet), d (doublet), dd
(doublet of doublets), t (triplet), dt (doublet of triplets), q (quartet), m (multiplet), bs (broad signal).
Mass spectra were recorded on an Applied Biosystem API-150 EX system spectrometer with an
ESI or an APCI interface. Compound 17 was synthesized according to literature methods.45
Compound 9 was purchased from Sigma-Aldrich and checked for purity by elemental analysis
(Calc. for C H NOS: C, 55.61; H, 3.33; N, 9.26. Found: C, 55.64; H, 3.21; N, 9.23). Purity of final
7
5
compounds was analyzed either on a ThermoQuest (Italia) FlashEA1112 Elemental Analyzer for C,
H, and N (analyses were within 0.4% of theoretical values) or by analytical high performance liquid
chromatography (HPLC) employing a Shimadzu HPLC gradient system (Shimadzu Corp., Kyoto,
Japan) equipped with two LC-10AD solvent delivery modules, a Rheodyne 7125 sample injector
and a SPD-10A UV-VIS detector. Prior to analysis, samples were prepared in MeOH at a final
concentration of 0.1 mg/mL. Chromatographic separation occurred on a Supelco Discovery C18
column (150 X 4.6 mm, 5 m particle size) by gradient elution. Flow rate was 1 mL/min and
injected volume was 10 L. Solvent A was methanol and Solvent B was water both additioned of
0
.1% v/v formic acid. Gradient conditions A: t(0 min): 55% A: 45% B; t(15 min): 95% A: 5% B;
ACS Paragon Plus Environment

Page 27 of 64
Journal of Medicinal Chemistry
1
2
3
t(17 min): 95% A: 5% B; returning to initial conditions after 1 min were employed for all
compounds. Purity results are presented as tR (min) and relative chemical purity (%). All tested
compounds were > 95% pure.
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
2
-Methylbenzo[d]isothiazol-3(2H)-one 10. To a stirred solution of benzo[d]isothiazol-3(2H)-one 9
(
61.2 mg, 0.39 mmol) in MeCN (2 ml), K CO (54.1 mg, 0.39 mmol) was added and the resulting
2
3
suspension is stirred for 10 min under nitrogen atmosphere. Iodomethane (100 L, 1.57 mmol) was
dropwise added over 5 min and the resulting mixture was stirred for 16 h. The volatile material is
removed by distillation under reduced pressure and the crude product is purified by FC (SiO , Et O)
2
2
to give 26.3 mg (40%) of the expected product 10 a as white waxy solid. Analytical data agree with
62
those reported. LC-UV tR purity: 4.36 min – 98.7%.
2-(6-Chlorohexaneyl)benzo[d]isothiazol-3(2H)-one 10a. To
a
stirred solution of
benzo[d]isothiazol-3(2H)-one 9 (100.4 mg 0.66 mmol) in CH CN (2 mL), K CO (91.7 mg, 0.67
3
2
3
mmol) was added and the resulting suspension was stirred for 10 min under nitrogen atmosphere. 1-
Chloro-6-iodohexane (1.0 mL, 6.6 mmol) was dropwise added over 5 min and the resulting mixture
was stirred for 18 h and then refluxed for 1 h. The solvent was removed by distillation under
reduced pressure and the crude was purified by FC (SiO , AcOEt/Hexane 50:50) to give 66.7 mg of
2
1
the expected product 10a as colorless oil (37%). H-NMR (300 MHz, CDCl )  8.03 (d, J= 7.9 Hz,
3
1
H); 7.49-7.42 (m, 1H); 7.40 (td, J = 8.0, 1.3 Hz, 1H); 3.90 (t, J = 7.1 Hz, 2H); 3.52 (t, J = 6.6 Hz,
2
H); 1.78 (m, 4H); 1.51 (m, 4H).
2
-(6-(Piperidin-1-yl)hexaneyl)benzo[d]isothiazol-3(2H)-one 11. In an amber glass vial with seal,
piperidine (58 μL, 0.6 mmol) was added to a stirred suspension of K CO (169.0 mg, 1.22 mmol),
2
3
KI (8.0 mg, 0.05 mmol) and 10a (66.4 mg, 0.25 mmol) in anhydrous DMF (1.0 mL). Upon
completion of addition, the reaction mixture was heated to 110 °C and stirred under nitrogen
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 64
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
atmosphere for 6 h. The volatiles were removed under reduced pressure and the crude reaction
mixture was diluted with AcOEt, the salts removed by filtration and the solvent evaporated under
1
reduced pressure. FC (SiO , DCM/MeOH 95:5) furnished 33.0 mg (41%) of 11 as colorless oil. H-
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
NMR (300 MHz, CDCl )  8.00 (d, J = 7.9 Hz, 1H); 7.61-7-52 (m, 2H); 7.38 (m, 1H); 3.87 (t, J =
3
7
.3 Hz, 2H); 2.35 (bs, 4H); 2.27 (t, J = 7.8 Hz, 2H); 1.75 (quintet, J = 7.1 Hz, 2H); 1.61-1.32 (m,
+
12H). MS-APCI: Calc. for C H N OS: 318.2, Found: 319.5 [M+H] . Anal. Calc. for C H N OS:
1
8
26
2
18 26
2
C, 67.88; H, 8.23; N, 8.80; Found: C, 67.49; H, 8.15; N, 8.63.
2
-(Chlorocarbonyl)phenyl hypochlorothioite 12. To a suspension of 2,2'-dithiosalicyclic acid
(2.02 g, 6.58 mmol) in toluene (10 mL) at 0 °C, thionyl chloride (1.65 g, 13.83 mmol) and a
catalytic amount of DMF (few drops) were added. The mixture was heated at 75 °C for 18 h and the
dark solution thus obtained was finally cooled to 0 °C. The crystals that separated were collected by
filtration, rinsed with cold hexane and air-dried to afford 1.60 g (70%) of 2,2'-dithiobisbenzoyl
chloride. The crude chloride was suspended at 0 °C in DCM (8.5 mL), and excess Cl was allowed
2
to bubble through the suspension. When a perfectly homogeneous solution is obtained, the stream
of chlorine was stopped and the solution of crude 12 was used as obtained.44
General method A. To a stirred solution of the amine (1.5 equiv) and anhydrous TEA (4.8 equiv)
in anhydrous DCM, a solution of 12 in anhydrous DCM was added dropwise over 2 min. The
reaction mixture was stirred at rt under nitrogen atmosphere for 16 h. The volatile components were
removed under reduced pressure, the residue was taken up with AcOEt and the resulting mixture
was washed sequentially with NaHCO saturated aqueous solution, 0.1M HCl aqueous solution and
3
brine. The organic phase was dried over Na SO and the solvent removed under reduced pressure.
2
4
2
-Phenylethylbenzo[d]isothiazol-3(2H)-one 13. According to method A, after purification by FC
(
SiO Hexane/AcOEt 70:30) and crystallization from n-hexane, 13 (16.1 mg, 11%) was obtained as
2
ACS Paragon Plus Environment

Page 29 of 64
Journal of Medicinal Chemistry
1
2
3
pale yellow crystals. Analytical data agree with those reported.63 LC-UV purity tR: 12.1 min –
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
9
9.8%.
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
2
-(2-(Naphthalen-1-yl)ethyl)benzo[d]isothiazol-3(2H)-one 14. According to method A, after
purification by FC (SiO Hexane/AcOEt 70:30) and crystallization from n-hexane, 14 (25 mg, 14%)
2
1
was obtained as pale yellow crystals. H-NMR (300 MHz, CDCl
3
) δ: 8.25 (d, J = 8.4 Hz, 1H); 8.07
(d, J = 7.8 Hz, 1H); 7.88 (d, J = 7.8 Hz, 1H); 7.78 (t, J = 4.5 Hz, 1H); 7.63-7.39 (m, 7H); 4.22 (t, J
=
7.8 Hz, 2H); 3.54 (t, J=7.8 Hz, 2H). MS-APCI: Calc. for C H NOS: 305.08, Found: 306.2
19 15
+
[M+H] . Anal. Calc. for C19
H15NOS: C, 74.72; H, 4.95; N, 4.59. Found: C, 74.66; H, 4.98; N, 4.61.
mp = 101-102 °C.
2
-(2-(Naphthalen-2-yl)ethyl)benzo[d]isothiazol-3(2H)-one 15. According to method A, after
purification by FC (SiO Hexane/AcOEt 70:30) and crystallization from n-hexane, 15 (28.7 mg,
2
1
1
1
6%) was obtained as pale yellow crystals. H-NMR (300 MHz, CDCl ) δ: H-NMR (300 MHz,
3
CDCl
3
) δ: 8.04 (d, J = 7.8 Hz, 1H); 7.76-7.83 (m, 3H); 7.71 (s, 1H); 7.58 (td, J = 8.2, 1.2 Hz, 1H);
7
.36-7.50 (m, 5H); 4.23 (t, J= 7.2 Hz, 2H); 3.24 (t, J = 7.5Hz, 2H). MS-APCI: Calc. for
+
C H NOS: 305.08, Found: 306.3 [M+H] . LC-UV purity tR: 14.8 min – 99.8%. mp = 114-115 °C.
1
9
15
2
-Cyclohexylbenzo[d]isothiazol-3(2H)-one 16. According to method A, after purification by FC
(SiO Hexane/AcOEt 70:30) and crystallization from n-hexane, 16 (31.2 mg, 30%) was obtained as
2
64
pale yellow crystals. Analytical data agree with those reported. Anal. Calc. for C H NOS: C,
1
6
15
6
6.92; H, 6.48; N, 6.00; Found: C, 66.68; H, 6.58; N, 5.82.
2
-Phenethylisothiazol-3(2H)-one 18. A solution of SO Cl in anhydrous 1,2-dichloroethane (0.75
2 2
M, 4.8 ml, 3.6 mmol) was added dropwise to a 0 °C stirring suspension of 3,3'-disulfanediylbis(N-
phenethylpropanamide) 17 (500.2 mg, 1.20 mmol) in anhydrous 1,2-dichloroethane (7.0 mL) over 2
ACS Paragon Plus Environment