The discovery of diazetidinyl diamides as potent and reversible inhibitors of monoacylglycerol lipase (MAGL)

Article abstract of DOI:10.1016/j.bmcl.2020.127198

Monoacylglycerol lipase (MAGL) has emerged as an attractive drug target because of its important role in regulating the endocannabinoid 2-arachidonoylglycerol (2-AG) and its hydrolysis product arachidonic acid (AA) in the brain. Herein, we report the discovery of a novel series of diazetidinyl diamide compounds 6 and 10 as potent reversible MAGL inhibitors. In addition to demonstrating potent MAGL inhibitory activity in the enzyme assay, the thiazole substituted diazetidinyl diamides 6d–l and compounds 10 were also effective at increasing 2-AG levels in a brain 2-AG accumulation assay in homogenized rat brain. Furthermore, selected compounds have been shown to achieve good brain penetration after oral administration in an animal study.

Full text of DOI:10.1016/j.bmcl.2020.127198

Bioorganic&MedicinalChemistryLetters30(2020)127198
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of diazetidinyl diamides as potent and reversible inhibitors of
monoacylglycerol lipase (MAGL)
Bin Zhu⁎, Peter J. Connolly, Sui-Po Zhang, Kristen M. Chevalier, Cynthia M. Milligan,
Christopher M. Flores, Mark J. Macielag
Janssen Research & Development, L.L.C., Welsh & McKean Roads, Spring House, PA 19477, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Monoacylglycerol lipase (MAGL) has emerged as an attractive drug target because of its important role in
regulating the endocannabinoid 2-arachidonoylglycerol (2-AG) and its hydrolysis product arachidonic acid (AA)
in the brain. Herein, we report the discovery of a novel series of diazetidinyl diamide compounds 6 and 10 as
potent reversible MAGL inhibitors. In addition to demonstrating potent MAGL inhibitory activity in the enzyme
assay, the thiazole substituted diazetidinyl diamides 6d–l and compounds 10 were also effective at increasing 2-
AG levels in a brain 2-AG accumulation assay in homogenized rat brain. Furthermore, selected compounds have
been shown to achieve good brain penetration after oral administration in an animal study.
Monoacylglycerol lipase (MAGL)
MAGL inhibition
Reversible MAGL inhibitor
2-Arachidonoylglycerol (2-AG)
Monoacylglycerol lipase (MAGL) is a serine hydrolase that is pri-
marily responsible for degrading 2-arachidonoylglycerol (2-AG) to
arachidonic acid (AA) and glycerol in the brain.1 2-AG is a prominent
endocannabinoid and an agonist for both cannabinoid receptor 1 (CB1)
and cannabinoid receptor 2 (CB2).2 AA, the 2-AG hydrolysis product by
MAGL, is a precursor for proinflammatory eicosanoids that can cause
neuroinflammation and lead to neurodegenerative diseases, such as
Alzheimer’s disease and Parkinson’s disease.3 Inactivation of MAGL has
been shown in animal models to elevate the level of 2-AG and reduce
the level of AA in the brain.4 Consequently, MAGL inhibition poten-
tially brings a dual benefit in treating diseases with an inflammatory
component, first by increasing 2-AG mediated cannabinoid receptor
activation and second by reducing the supply of AA-derived proin-
flammatory eicosanoids.5 Furthermore, MAGL has been implicated in
certain cancer cell growth; therefore, MAGL inhibitors may also be
useful as anti-cancer agents.5,6 As a result, MAGL has gathered in-
creasing attention as an attractive drug target, and a number of MAGL
inhibitors have been reported. The majority of those reported in-
hibitors, however, are irreversible and form a covalent bond with
MAGL;5c,7 and studies have shown that chronic inhibition of MAGL by
an irreversible MAGL inhibitor or genetic deletion of MAGL leads to
desensitization of CB1, which eventually impairs the beneficial 2-AG
mediated effects that are CB1-dependent.4a,4b A potential approach to
alleviate such drawbacks associated with irreversible inhibition of
MAGL is to identify reversible MAGL inhibitors for therapeutic use.5c,8
We have previously reported a potent reversible MAGL inhibitor
(compound 1, Fig. 1).9 Structural optimization based on compound 1 to
improve metabolic stability led to the discovery of a novel series of
diazetidinyl diamides (Fig. 1) as potent reversible inhibitors of MAGL.
The synthesis of diazetidinyl diamide compounds 6 is shown in
Scheme 1. 1-Boc-3-amino-azetidine 2 was reacted with carboxylic acid
(R1CO2H) or acid chloride (R1COCl) under standard amide formation
conditions to give compound 3. Removal of the Boc protecting group
under acidic conditions (CF3CO2H, CH2Cl2) followed by reductive
amination with 1-Boc-3-azetidinone 4 [Na(OAc)3BH, acetic acid, 1,2-
dichloroethane] led to compound 5. The Boc protecting group of 5 was
removed (CF3CO2H, CH2Cl2), and the resulting product was reacted
with a second carboxylic acid (R2CO2H) or acid chloride (R2COCl) to
give the final product diazetidinyl diamide 6.
Representative diazetidinyl diamide compounds 6 are summarized
in Table 1. In vitro MAGL inhibitory activity of the compounds was
measured in an enzyme assay using purified wild-type human MAGL or
mutant human MAGL,10 which is a more soluble form of the wild-type
enzyme with comparable catalytic activity.9 In addition, a brain 2-AG
accumulation assay was established to assess a given compound’s ef-
fectiveness in elevating 2-AG levels in rat brain tissue, a consequence of
MAGL inhibition.11 In this assay, 2-AG levels in homogenized rat brain
incubated with test compound were measured and compared to those
incubated with the vehicle. The activity of the test compound is ex-
pressed as percent brain 2-AG accumulation, which is indicative of a
compound’s effectiveness at inhibiting MAGL in brain. Higher brain 2-
AG accumulation (%) indicates a more effective inhibitor.
⁎ Corresponding author.
E-mail address: bzhu@its.jnj.com (B. Zhu).
https://doi.org/10.1016/j.bmcl.2020.127198
Received 20 November 2019; Received in revised form 10 April 2020; Accepted 12 April 2020
Availableonline18April2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

B. Zhu, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127198
Fig. 1. Design of diazetidinyl diamides as re-
versible MAGL inhibitors.
Table 1
MAGL inhibitory activity and 2-AG accumulation of selected compounds
6.
Cpd R1
R2
MAGL
Brain 2-AG
IC50 (nM)
accumulation (%)
(Cpd conc.)
1
10a
N/A
N/A
Scheme 1. a) R1CO2H, HATU, Et3N, CH2C12; or R1COC1, Et3N, CH2C12; b)
TFA, CH2C12; c) 4, Na(OAc)3BH, HOAc, 1,2-dichloroethane; d) R2CO2H, HATU,
Et3N, CH2C12; or R2COC1, Et3N, CH2C12.
6a
270a
6b
6c
939a
N/A
As illustrated by the data in Table 1, thiazole as the R1 group led to
compounds with much higher potency in the MAGL enzyme assay than
the corresponding analogs where the R1 group is phenyl (6d, 6e, 6h vs
6a–c). The 2-thiazole and 4-thiazole R1 groups were not substantially
different (6f, 6g, 6i vs 6j–l). When the R1 group was 2-thiazole, analogs
with fused heteroaromatic R2 groups such as 2-phenyl benzoxazole
(6e), N-phenyl-indole (6g), and benzothiophene (6h, 6i), as well as
analogs with bicyclic aromatic R2 groups such as benzylphenyl (6d) and
3′-CF3-biphenyl (6f) exhibit good potency in the MAGL enzyme assay
(IC50) and in the brain 2-AG accumulation assay (Brain 2-AG Accu-
mulation%).
458a
N/A
6d
6e
15a
48a
659% (10 μM)
486% (10 μM)
355% (1 μM)
6f
7a
921% (10 μM)
347% (1 μM)
6g
12b
338% (1 μM)
The crystal structure of compound 6d binding with mutant human
MAGL (hMGL 1-303 K36A, L169S, L176S) was obtained (Fig. 2), which
confirmed noncovalent binding of compound 6d to MAGL.12 Com-
pound 6d occupies the same binding pocket in MAGL as compound 1,
which is located between helices α4, α5, α6 and α7.9 Similar as com-
pound 1, the azetidine-amide carbonyl of compound 6d points into the
oxyanion hole of MAGL and forms a hydrogen bond with the backbone
amide NH of Met123 that is adjacent to the catalytic Ser122. In addi-
tion, the thiazole-amide carbonyl of compound 6d makes a hydrogen
bond to the sidechain guanidine group of Arg57, and there is a π-π
stacking interaction between the thiazole-amide and Tyr194 (Fig. 3).
Because compounds 6a–c (R1 = phenyl) showed weak MAGL in-
hibitory activity and to further explore heterocyclic/heteroaromatic
substitution at R1, the related cyclic amide analogs 10 were synthesized
(Scheme 2) and evaluated. Reductive amination [Na(OAc)3BH, HOAc,
1,2-dichloroethane] of methyl 2-formylbenzoate 7 with 1-Boc-3-amino-
azetidine 2 followed by spontaneous lactam formation gave the cy-
clized amide intermediate 8. Deprotection under acidic conditions
(CF3CO2H, CH2Cl2) removed the azetidine Boc group, and a second
reductive amination with 1-Boc-3-azetidinone 4 [Na(OAc)3BH, acetic
acid, 1,2-dichloroethane] led to compound 9. Removal of the Boc
protecting group of 9 (CF3CO2H, CH2Cl2) followed by an amide for-
mation reaction with carboxylic acid (R2CO2H) or acid chloride
(R2COCl) resulted in the final product diazetidinyl diamide 10.
Results for a selection of compounds 10 are summarized in Table 2.
The data indicate compounds 10 achieve a significant improvement of
6h
6i
28a
611% (10 μM)
460% (1 μM)
< 5a
6j
6a
620% (1 μM)
220% (10 μM)
6k
15a
6l
< 5a
477% (1 μM)
a
b
Mutant MAGL.
Wild-type MAGL; N/A: Data not available.
were tested against fatty acid amide hydrolase (FAAH), and all three
compounds showed no significant inhibitory activity (FAAH
IC50 > 10 μM). Additionally, based on their in vitro potency and ADME
properties, several compounds were selected to be evaluated in a rat
tissue distribution study (10 mg/Kg, po) for their ability to penetrate
the blood–brain barrier (BBB). Blood and brain samples were collected
at the 1 h time point and analyzed. Among the 4 compounds in Table 3,
compound 10f achieved the highest brain concentration (1.20 nmol/g),
while compound 6f showed the highest brain/plasma ratio (0.85).
In summary, a novel series of diazetidinyl diamide compounds were
discovered as potent reversible MAGL inhibitors. The thiazole
MAGL inhibitory activity compared to compounds
6 of which
R1 = phenyl (10a vs 6b). The compounds 10 illustrated in Table 2
demonstrate similar potency to compounds 6 with R1 = 2-thiazole in
Table 1, considering both MAGL enzyme activity and brain 2-AG ac-
cumulation (10a, 10c, 10e vs 6e, 6g, 6i).
To assess their selectivity for MAGL, compounds 6e, 6f and 10e
2

B. Zhu, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127198
Scheme 2. a) Na(OAc)3BH, HOAc, 1,2-dichloroethane; b) TFA, CH2C12; c) 4,
Na(OAc)3BH, HOAc, 1,2-dichloroethane; d) R2CO2H, HATU, Et3N, CH2C12; or
R2COC1, Et3N, CH2C12.
Table 2
MAGL inhibitory activity and 2-AG accumulation of selected compounds
10.
Cpd
R2
MAGL IC50
(nM)
Brain 2-AG Accumulation (%)
(Cpd conc.)
Fig. 2. Overview of the cocrystal structure of compound 6 d (green) and mutant
human MAGL.
10a
31a
435% (10 μM)
152% (1 μM)
10b
10c
15a
11b
725% (10 μM)
204% (1 μM)
332% (1 μM)
10d
10e
19a
409% (10 μM)
192% (1 μM)
449% (1 μM)
< 5a
10f
18a
676% (10 μM)
253% (1 μM)
a
b
Mutant MAGL.
Wild-type MAGL.
Fig. 3. A close-up view of the compound 6d (green) binding site to MAGL.
Table 3
Plasma and brain concentrations of selected compounds.
substituted compounds 6d–l (R1 = thiazole) exhibited high MAGL in-
hibitory potency. Structural modification of the less potent phenyl
substituted compounds 6a–c (R1 = phenyl) led to a series of cyclic
amide analogs 10 with much more improved MAGL inhibitory activity.
Selected compounds also demonstrated good brain penetration in rat
after an oral dose. Such compounds may serve as tools for future re-
search of MAGL inhibition as a therapeutic approach to treat disorders
including pain, inflammation, and depression.
Cpd
Plasma conc. (μM)
Brain conc. (nmol/g)
B/P ratio
6e
6f
2.17
0.81
1.08
4.76
0.84
0.67
0.35
1.20
0.39
0.85
0.33
0.25
6j
10f
References
1. (a) Dinh TP, Carpenter D, Leslie FM, et al. Proc Natl Acad Sci USA.
2002;99:10819–10824;
Declaration of Competing Interest
(b) Dinh TP, Freund TF, Piomelli D. Chem Phys Lipids. 2002;121:149–158;
(c) Blankman JL, Simon GM, Cravatt BF. Chem Biol. 2007;14:1347–1356.
2. (a) Mechoulam R, Ben-Shabat S, Hanus L, et al. Biochem Pharmacol. 1995;50:83–90;
(b) Hillard CJ. Prostaglandins Other Lipid Mediators. 2000;61:3–18;
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
(c) Gonsiorek W, Lunn C, Fan X, et al. Mol Pharmacol. 2000;57:1045–1050.
3. Nomura DK, Morrison BE, Blankman JL, et al. Science. 2011;334:809–813.
4. (a) Schlosburg JEl, Blankman JL, Long JZ, et al. Nat Neurosci. 2010;13:1113–1119;
3

B. Zhu, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127198
(b) Chanda PK, Gao Y, Mark L, et al. Mol Pharmacol. 2010;78:996–1003;
(50 nL) into the assay plate using a Cartesian Hummingbird prior to adding 4MU-B
(25 μL of 1.2X solution to a final concentration of 10 μM) followed by enzyme (5 μL
of a 6X solution to a final concentration of 5 nM) to initiate the reaction. Final
compound concentrations ranged from 17 to 0.0003 μM. The fluorescence change
due to 4MU-B cleavage was monitored with excitation and emission wavelengths of
335 and 440 nm, respectively, and a bandwidth of 10 nm (Safire2, Tecan) at 37°C for
5 min. The IC50 values for the test compounds were determined using Excel from a
fit of the equation to the concentration-response plot of the fractional activity as a
function of inhibitor concentration.
(c) Long JZ, Li W, Burston JJ, et al. Nat Chem Biol. 2009;5:37–44.
5. (a) For recent reviews, see Tuo W, Lelue-Chavain N, Spencer J, et al. J Med Chem.
2016;60:4–46;
(b) Grabner GF, Zimmermann R, Schicho R, et al. Pharmacol Therap.
2017;175:35–46;
(c) Gil-Ordonez A, Martin-Fontecha M, Ortega-Gutierrez S, et al. Biochem Pharmacol.
2018;157:18–32.
6. Nomura DK, Lombardi DP, Chang JW, et al. Chem Biol. 2011;18:846–856.
7. (a) Brindisi M, Maramai S, Gemma S, et al. J. Med Chem. 2016;59:2612–2632;
(b) Cisar JS, Weber OD, Clapper JR, et al. J Med Chem. 2018;61:9062–9084.
8. (a) King AR, Dotsey EY, Lodola A, et al. Chem Biol. 2009;16:1045–1052;
(b) Hernandez-Torres G, Cipriano M, Heden E, et al. Angew Chem Int Ed Engl.
2014;53:13765–13770;
11. One gram of rat brain was homogenized using a Polytron homogenizer (Brinkmann,
PT300) in 10 mL of 20 mM HEPES buffer (pH = 7.4), containing 125 mM NaCl, 1
mM EDTA, 5 mM KCl, and 20 mM glucose. Test compounds (1 μM) or vehicle were
pre-incubated with rat brain homogenate (50 mg). After a 15-min incubation time at
37°C, CaCl2 (final concentration = 10 mM) was added and then incubated for 15
min at 37°C in a total volume of 5 mL. The reaction was stopped with 6 mL organic
solvent extraction solution of 2:1 chloroform/methanol. The concentration of 2-AG
in the organic phase was measured by a HPLC/MS method, and Brain 2-AG
Accumulation% of the compound was derived from the following formula.
12. MAGL construct (hMAGL 1-303, K36A, L169S, L176S mutant) in 50 mM HEPES
buffer (pH 7.5) with 200 mM NaCl, 2% glycerol, 2 mM EDTA, and 2 mM dithio-
threitol, was mixed with 2-fold excess of (3-iodo-4-methylphenyl)(3-(4-(pyrimidin-2-
yl)piperazin-1-yl)azetidin-1-yl)methanone and incubated overnight at 4°C and was
then concentrated to 5-6mg/mL for crystallization. Crystals were obtained by mixing
1 μL of complexed protein with 0.2 μL seed stock and 0.5 μL of 8–13% PEGMME
5000, 100 mM Na-Citrate (pH 5.5), 0.2% glucopyranoside and equilibrated against 1
mL reservoir of 6–11% PEGMME 5000, 100 mM Na-Citrate pH 5.5, 0.2% gluco-
pyranoside in a hanging drop vapor diffusion experiment at room temperature. These
co-crystals were then used for displacement soaks with compound 6d (5 mM, over-
night soak). The resulting crystals were harvested, transferred briefly into cryopro-
tectant containing 16% PEGMME 5000, 100 mM Na-MES (pH 6.0), and 25% gly-
cerol, and flash frozen in liquid nitrogen. A complete dataset was collected for both
on a Rigaku M007HF generator at 100°K. Data was processed in d*trek and structure
refined in PHENIX.
(c) Granchi C, Rizzolio F, Bordoni V, et al. J Enzyme Inhib Med Chem.
2016;31:137–146;
(d) Granchi C, Caligiuri I, Bertelli E, et al. J Enzyme Inhib Med Chem.
2017;32:1240–1252;
(e) Tuccinardi T, Granchi C, Rizzolio F, et al. Bioorg Med Chem. 2014;22:3285–3291;
(f) Granchi C, Rizzolio F, Palazzolo S, et al. J Med Chem. 2016;59:10299–10314;
(g) Aghazadeh Tabrizi M, Baraldi PG, Baraldi S, et al. J Med Chem.
2018;61:1340–1354;
(h) Aida J, Fushimi M, Kusumoto T, et al. J Med Chem. 2018;61:9205–9217;
(i) Dato FM, Neudorfl J, Gutschow M, et al. Bioorg Chem. 2020;94;
(j) Wyatt RM, Fraser I, Welty N, et al. Journal of Pharmacol Exp Ther.
2020;373:339–353.
9. Schalk-Hihi C, Schubert C, Alexander R, et al. Protein Sci. 2011;20:670–683.
10. All rate-based assays were performed in black 384-well polypropylene polymerase
chain reaction (“PCR”) microplates (Abgene) in a total volume of 30 μL. Substrate 4-
methylumbelliferyl butyrate (4MU-B; Sigma) and purified mutant MAGL (mut-MGLL
11-313 L179S L186S) or wild type MGL (wt-MGLL 6H-11-313) were diluted sepa-
rately into 20 mM 1,4-piperazinediethanesulfonic acid (“PIPES”) buffer (pH = 7.0),
containing 150 mM NaCl and 0.001% Tween 20. Test compounds were pre-dispensed
4