Discovery of ACH-000143: A Novel Potent and Peripherally Preferred Melatonin Receptor Agonist that Reduces Liver Triglycerides and Steatosis in Diet-Induced Obese Rats

pubs.acs.org/jmc

Article

Discovery of ACH-000143: A Novel Potent and Peripherally

Preferred Melatonin Receptor Agonist that Reduces Liver

Triglycerides and Steatosis in Diet-Induced Obese Rats

Marcos Antonio Ferreira, Jr.,^#Hatylas Azevedo,*^,#Alessandra Mascarello,^#Natanael Dante Segretti,

Elisa Russo, Valter Russo, and Cristiano Ruch Werneck Guimaraes

̃

Cite This: J. Med. Chem. 2021, 64, 1904 −1929

Read Online

sı

*

Metrics & More

Article Recommendations

Supporting Information

ACCESS

ABSTRACT: The modulation of melatonin signaling in peripheral tissues holds promise for treating metabolic diseases like obesity,

diabetes, and nonalcoholic steatohepatitis. Here, several benzimidazole derivatives have been identiﬁed as novel agonists of the

melatonin receptors MT1 and MT2. The lead compounds 10b, 15a, and 19a demonstrated subnanomolar potency at MT1/MT2

receptors, high oral bioavailability in rodents, peripherally preferred exposure, and excellent selectivity in a broad panel of targets.

Two-month oral administration of 10b in high-fat diet rats led to a reduction in body weight gain similar to dapagliﬂozin with

superior results on hepatic steatosis and triglyceride levels. An early toxicological assessment indicated that 10b (also codiﬁed as

ACH-000143) was devoid of hERG binding, genotoxicity, and behavioral alterations at doses up to 100 mg/kg p.o., supporting

further investigation of this compound as a drug candidate.

ameliorates methionine- and choline-deﬁcient diet-induced

nonalcoholic steatohepatitis (NASH) in rats.¹⁸Additionally,

melatonin displays synergistic actions with metformin on

improving adiposity and insulin sensitivity in DIO rats exhibiting

circadian disruption.¹⁹These ﬁndings suggest that drugs that

regulate circadian disruption linked to metabolic dysfunctions

could have a positive impact in the treatment of obesity,

diabetes, and liver metabolic diseases.

The physiological eﬀects of melatonin occur through diﬀerent

molecular pathways, but the best-characterized signaling

involves the activation of G protein-coupled receptors

(GPCRs), named the MT1 and MT2 melatonin receptors.²⁰

INTRODUCTION

■

In recent years, much attention has been devoted to under-

standing the link between the desynchronization of central and

peripheral clocks and the development of metabolic syn-

drome.^1,2Animals with mutations in clock genes are obese

and exhibit hyperglycemia, hyperlipidemia, and liver steato-

sis.³⁻⁶Moreover, genetic polymorphisms in circadian genes

have been associated with metabolic alterations.^7,8Several

studies in shift workers^9,10and in diabetic and obese

patients¹¹⁻¹³have also highlighted the connection between

circadian rhythm and metabolic dysregulation.

Melatonin (N-acetyl-5-methoxytryptamine) is a natural

hormone that inﬂuences the circadian regulation of glucose

and insulin levels, synchronizing the metabolism with daily

feeding and fasting cycle.¹⁴Genetic mutations in melatonin

receptors and altered melatonin signaling have been related to

increased fasting plasma glucose levels, impaired insulin

secretion, and higher risk of type 2 diabetes.^15,16Melatonin

administration improves liver steatosis, inﬂammatory markers,

and lipid levels in high fat diet-induced obese (DIO) mice¹⁷and

Received: April 23, 2020

Published: February 8, 2021

American Chemical Society

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

1904

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 1. Synthesis of Final Compounds 4, 6, 7, and 8a−j

a

(a) NH₂(CH₂)₂NH₂, CuBr₂, 60 °C, 2 h; (b) Ac₂O or R³-COCl, EtOH, 80 °C, 1 h; (c) CH₂O₂, THF, Pd/C (10%), H₂(35 bar), 60 °C, 8 h; (d)

Zn powder, NH₄HCO₂, MeOH, 45 °C or Pd/C (10%), H₂(50 psi), MeOH, 30 °C, 4 h; (e) BTC, TEA, THF, 0 °C; (f) tetramethoxymethane or

tetraethoxymethane, AcOH, 80 °C, 30 min; (g) thiourea, EtOH, 80 °C; (h) K₂CO₃, CH₃I, acetone.

Besides the action on these GPCRs, melatonin also has anti-

oxidative properties and binds to intracellular proteins, such as

calmodulin and nuclear receptors.^21,22This pleiotropic action

results in several positive eﬀects, such as reduced hyperglycemia,

insulin resistance, oxidative stress, and inﬂammation.²³How-

ever, it has been diﬃcult to deﬁne which metabolic eﬀects of

melatonin rely on its action at central or peripheral MT1/MT2

receptors due to the ubiquitous presence of these receptors and

melatonin itself in multiple tissues, like in the brain, gastro-

intestinal tract, liver, kidney, and pancreas.^24,25In addition, the

lipophilic nature and high brain-to-plasma ratio of synthetic

MT1/MT2 ligands^26,27do not provide optimal pharmacological

tools to investigate the speciﬁc peripheral contributions of

melatonin signaling. Therefore, the discovery of potent and less

lipophilic MT1/MT2 agonists, peripherally preferred, is

required to explore the metabolic eﬀects following the peripheral

activation of MT1 and/or MT2.

The study of synthetic MT1/MT2 agonists has initially

explored replacing the methoxy and N-acetyl groups of

melatonin by larger alkoxy and N-acyl substituents.²⁸Potent

compounds have also been obtained by conformational

restriction of the amide side chain and its incorporation into

additional rings.²⁹Further, the indole nucleus has been replaced

by other aromatic moieties, such as indane, naphthalene, and

benzofuran.³⁰This approach resulted in potent, lipophilic, non-

selective MT1/MT2 agonists with high brain exposure, such as

ramelteon ((S)-N-[2-(1,6,7,8-tetrahydro-2H-indeno-[5,4-b]-

furan-8-yl)ethyl]propionamide), agomelatine (N-[2-(7-me-

thoxy-1-naphthyl)ethyl]acetamide), and tasimelteon (N-

[[(1R,2R)-2-(2,3-dihydro-1-benzofuran-4-yl)cyclopropyl]-

methyl]propenamide). These drugs have been approved for the

treatment of insomnia, depression, and non-24 h sleep−wake

disorder.

Previously described melatonergic ligands with benzimida-

zole scaﬀold have failed to achieve similar potency level as

melatonin at the MT1/MT2 receptors.^31,32The direct

melatonin analogue obtained by replacement of the indole

ring with benzimidazole has a much lower aﬃnity for melatonin

receptors than melatonin (K_i= 1.70 μM versus 0.09 nM) in

binding studies using ovine pars tuberalis membrane homoge-

nates.³¹On the other hand, a tricyclic dihydrofuran

benzimidazole derivative with a phenyl substituent at the 2-

position of the benzimidazole ring displays improved binding

aﬃnities for MT1 and MT2 (K_i= 20 and 4.5 nM,

respectively).³³This suggests that substitutions around the

benzimidazole ring and possibly modulation of its basicity/

lipophilicity could compensate for the loss in binding aﬃnity

caused by replacement of the melatonin’s indole group. We view

the lower lipophilic nature of the benzimidazole core as strategic

to achieve peripheral over central distribution as well as

improved pharmacokinetics (PK) over marketed MT1/MT2

agonists.

1905

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Scheme 2. Synthesis of Final Compounds 9, 10a, and 10b

a

(a) NCS, i-PrOH, 90 °C, 16 h; (b) NaOH, MeOH, 80 °C, 8 h.

a

Scheme 3. Synthesis of Final Compounds 14a, 14b, 15a, and 15b

a

(a) NH₂(CH₂)₃NH₂, CuBr₂, 60 °C, 2 h; (b) Ac₂O, EtOH, 80 °C, 1 h; (c) Zn powder, NH₄HCO₂, MeOH, 45 °C or Pd/C (10%), H₂(50 psi),

MeOH, 30 °C, 4 h; (d) tetramethoxymethane or tetraethoxymethane, AcOH, 80 °C, 30 min; (e) NCS, i-PrOH, 90 °C, 16 h.

a

Scheme 4. Synthesis of Compounds 19a and 19b

a

(a) Tetramethoxymethane or tetraethoxymethane, AcOH, 80 °C, 30 min; (b) TFA, DCM, 20 °C, 6 h; (c) Ac₂O, EtOH, 80 °C, 1 h.

In the present work, several benzimidazole derivatives have

been synthesized and tested in vitro to determine their activity at

the MT1 and MT2 receptors, physicochemical and ADME

properties, as well as selectivity across a panel of several

receptors, enzymes, and transporters. Moreover, in vivo PK

studies have been performed in rodents to characterize the

plasma and brain PK proﬁles and guide the selection of

compounds for in vivo eﬃcacy studies. Speciﬁcally, the

metabolic eﬀects following repeated administration of three

novel benzimidazole derivatives with potent agonism at MT1/

MT2 have been investigated on plasma, liver, and muscle tissues

of DIO rats.

and MT2 receptors,^34,35ligands enter the melatonin binding site

through a membrane-buried lateral channel. While this channel

is easily accessible for the non-ionizable melatonin, the reduced

lipophilicity and/or basic nature of the benzimidazole core

could negatively impact its diﬀusion across lipophilic mem-

branes as well as binding to MT1/MT2 receptors.³⁶However,

the lipophilicity and pK_aof the benzimidazole derivatives could

be ﬁne-tuned by the introduction of substituents, such as at the

C2 and C5 positions, to ﬁnd the optimal balance of properties.

To probe this, the pK_avalues for the conjugated acids of the

model compounds 1-methylbenzimidazole and its 2-methoxy

derivative were estimated using a density functional theory

(DFT)-based method.³⁷The results indicate that the 2-methoxy

substitution reduces the calculated pK_aby a full log unit (6.0

versus 5.0, Table S1), suggesting that the electron-withdrawing

inductive eﬀect due to the substitution is more signiﬁcant than

the electron-donating eﬀect by resonance. The C2-substitution

is also expected to enhance binding to MT1/MT2 by occupying

a hydrophobic subpocket present in both receptors.

RESULTS

■

Design and Synthesis of Novel Benzimidazole De-

rivatives. The direct melatonin analogue obtained by

replacement of the indole ring with benzimidazole has been

reported to have a 15,000-fold lower aﬃnity for melatonin

receptors.^31,32As revealed by recent crystal structures of MT1

1906

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 1. In Vitro Pharmacology at MT1 And MT2 Receptors (Binding and Functional Assays) of Compounds 4, 6, 7, 8a, 8d, 8j,

a

19a, and 19b

a

The assays were conducted using CHO cells expressing human MT1 or MT2. For the binding assay, compounds were incubated with the cell

membranes and the radioligand ([¹²⁵I]2-iodomelatonin), and the readout was scintillation counting. Binding results are expressed as radioligand

binding inhibition or inhibition constant (Ki). Functional assays were performed in the agonist mode with cell impedance and cAMP as readouts

for MT1 and MT2, respectively. Compounds were tested in a log or semi-log dilution within 0.01 nM to 100 μM. All compounds were full agonists.

EC₅₀: half-maximal eﬀective concentration. NT: not tested. Experiments were performed once in duplicate. Data are presented as mean SEM.

We synthesized a series of novel benzimidazoles using

convenient routes (Schemes 1, 2, 3, and 4). Scheme 1 illustrates

the synthesis of compounds 4, 6, 7, and 8a−j using 1-ﬂuoro-4-

methoxy-2-nitrobenzene (or 2-ﬂuoro-4-methoxy-1-nitroben-

zene) as the starting material, which was combined with

ethylenediamine and CuBr₂to aﬀord the nitroaniline deriva-

tive.³⁸Acylation of this intermediate with acetic anhydride or

acyl halides produced the desired acylated compounds 3a−i in

good yields. Nitro reduction followed by cyclization with formic

acid gave exclusively the desired compound 4, as described by

Depreux et al.³⁸Alternatively, the nitro group was reduced with

Zn powder/NH₄HCO₂in MeOH or Pd/C (10%) in MeOH to

provide the desired phenylenediamines 5a−i; 5a was then

treated with bis(trichloromethyl)carbonate (BTC)^39,40to give

the 5-membered cyclic urea 6. Tetramethoxymethane or

tetraethoxymethane in acetic acid were used to react with the

phenylenediamine 5a−i aforementioned to give the ﬁnal

compounds 8a−j. Finally, compound 7 was obtained after

treatment of the intermediate phenylenediamine 5a with

thiourea in EtOH followed by an alkylation reaction with

CH₃I in acetone.

Compounds 9, 10a, and 10b were prepared from compounds

8a and 8d as described in Scheme 2. Hydrolysis of compound 8a

with sodium hydroxide followed by the HCl/EtOH treatment

aﬀorded the desired compound 9. Chlorination of 8a and 8d

using N-chlorosuccinimide in isopropyl alcohol gave the desired

compounds 10a and 10b, respectively. To allow the exploration

of additional derivatives with three carbon atoms in the

1907

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Table 2. In Vitro Pharmacology at MT1 and MT2 Receptors (Binding and Functional Assays) of Compounds 8a−c, 8e−i, and 9

a

Radioligand binding assays were conducted using CHO cells expressing human recombinant MT1 or MT2 receptors. Compounds were incubated

with the cells and the radioligand ([¹²⁵I]2-iodomelatonin), and the assay readout was measured via scintillation counting. Binding results are

expressed as the radioligand binding inhibition or inhibition constant (K_i). Cellular functional assays were performed in the agonist mode using

CHO cells expressing human MT1 or MT2 receptors. The assay readout was cell impedance for MT1, measured by cellular dielectric spectroscopy,

and cAMP for MT2, measured by a ﬂuorometric method. Compounds were assessed in a log or semi-log dilution within the concentration range

0.01 nM to 100 μM. All compounds were full agonists. EC₅₀: half-maximal eﬀective concentration. NT: not tested. Experiments were performed

once in duplicate. Values are reported as the mean SEM.

alkylamide tail, compounds 14a, 14b, 15a, and 15b were

synthesized following the route previously described for

compound 8a and using 1,3-diaminopropane to aﬀord the

nitroaniline intermediate, as shown in Scheme 3.

As illustrated in Scheme 4, the intermediate 16, compound

tert-butyl{2-[(5-amino-2,3-dihydro-1-benzofuran-4-yl), synthe-

sized as previously described by Koike et al.,³³was treated with

tetramethoxymethane or tetraethoxymethane to give the

benzimidazole core. Finally, the Boc group was deprotected

followed by a reaction with acetic anhydride to aﬀord

compounds 19a and 19b.

The compounds were obtained with yields ranging from 32 to

80%, and their chemical structures were characterized by the

combined analysis of NMR and MS data (see the Experimental

Section). In addition, NMR spectra of the ﬁnal compounds are

reported in Figures S7 −S40.

Compounds 4, 6, 7, 8a, 8d, 8j, 19a, and 19b have been used to

test the impact of substituents on the benzimidazole core (Table

1). The results were compared with the in vitro activity of

melatonin, a non-selective MT1/MT2 full agonist (K_i= 0.18

and 0.12 nM; EC₅₀= 0.11 and 0.07 nM at MT1 and MT2,

respectively). Remarkably, the 2-ethoxybenzimidazole deriva-

tive 8a had a proﬁle similar to melatonin, exhibiting only 2- to 8-

fold lower MT1/MT2 aﬃnity and agonist potency (K_i= 0.42

and 0.93 nM; EC₅₀= 0.33 and 0.26 nM at MT1 and MT2,

respectively). Moving the methoxy group from position 6 to

position 5 resulted in compound 8j, which displayed lower

binding aﬃnity at both receptors in comparison to compound

8a (60.7% vs 98.9% for MT1 and 36.5% vs 98.7% for MT2). In

contrast, the cyclization of the 6-methoxy substituent into a

tetrahydrofuran ring (19a and 19b) was tolerated in MT1 and

increased the binding aﬃnity and potency at MT2 up to 5-fold

when compared with 8a.

We also explored substitutions at the 2-position of the

benzimidazole core by comparing the potency of the 2-ethoxy

substitution (8a) versus 2-methoxy (8d), 2-oxo (6), 2-

thiomethyl (7), and unsubstituted (4) analogues. While the

unsubstituted and 2-oxo analogues, 4 and 6, respectively,

showed considerably reduced MT1/MT2 binding (49−67% at

1 μM) than the 2-ethoxy derivative 8a (99% at 1 μM), the 2-

thiomethyl (7) and 2-methoxy (8d) substituents were well

tolerated resulting in similar aﬃnity and agonist potency (Table

1).

Structure−Activity Relationship (SAR) of Novel Benzi-

midazole Derivatives at the MT1 and MT2 Receptors.

Compounds were tested in vitro for their binding aﬃnity and

potency at the MT1/MT2 receptors using radioligand and

agonist functional assays. Competitive radioligand binding

assays were conducted using membranes from CHO cells

recombinantly expressing human MT1 or MT2, the radioligand

[

¹²⁵I]2-iodomelatonin and scintillation counting as the

quantiﬁcation method. Binding results are expressed as the

radioligand binding inhibition or inhibition constant (K_i).

Cellular functional assays were performed in the agonist mode

using CHO cells expressing human MT1 or MT2. In the

screening phase, the assay readouts were cell impedance for

MT1, measured by cellular dielectric spectroscopy, and cAMP

for MT2, measured by a ﬂuorometric method. Twenty-two

compounds were initially screened in MT1 and MT2 for their

aﬃnities at 1 μM (% of binding), and the K_i(inhibition

constant) and EC₅₀(functional assay) values were determined

for compounds with most favorable aﬃnities.

Substitutions were also explored at the side chain amide group

(Table 2). Small alkyl groups, such as ethyl (8b), n-propyl (8c),

and cyclopropyl (8e) were tolerated (8b and 8e) or resulted in

slightly higher aﬃnity and potency (8c) at MT1 and MT2, when

compared to the acetamide (8a). In contrast, replacement with

bulkier groups like cyclobutane (8f), cyclopentane (8g),

cyclohexane (8h), and oxopyran (8i) were detrimental, and

the substitution of the amide group with an amine resulted in

loss of activity (9). The number of carbon atoms was also

1908

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Table 3. In Vitro Pharmacology at MT1 and MT2 Receptors (Binding and Functional Assays) of Compounds 8a, 8d, 10a, 10b,

a

14a, 14b, 15a, and 15b

Binding (%) at 1 μM

Binding K_i(nM)

Agonism EC₅₀(nM)

Compound

n

R1

R2

MT1

MT2

MT1

MT2

MT1

MT2

8a

8d

2

3

−H

−Cl

−H

−Cl

−CH₂CH₃

−CH₃

−CH₂CH₃

−CH₃

−CH₂CH₃

−CH₃

−CH₂CH₃

−CH₃

98.9 0.1

100.2 0.9

100.1 0.8

100.3 0.3

99.4 0.6

98.3 0.4

95.7 0.2

81.8 0.3

98.7 0.4

98.3 0.1

97.5 0.4

99.4 0.4

95.7 1.5

97.6 0.2

99.5 0.3

98.2 0.3

0.42 0.20

1.20 0.08

0.88 0.01

2.1 0.1

1.10 0.14

2.20 0.01

13.0 1.4

NT

0.93 0.10

3.1 0.1

0.33 0.01

0.20 0.03

0.14 0.01

0.25 0.04

0.18 0.03

1.92 0.04

2.13 0.49

NT

0.26 0.10

0.52 0.13

0.27 0.01

0.86 0.03

0.35 0.08

2.01 0.44

0.30 0.03

NT

10a

10b

14a

14b

15a

15b

0.93 0.08

0.80 0.02

4.50 0.08

11.0 0.1

1.60 0.30

NT

a

Radioligand binding assays were conducted using CHO cells expressing human recombinant MT1 or MT2 receptors. Compounds were incubated

with the cells and the radioligand ([¹²⁵I]2-iodomelatonin), and the assay readout was measured via scintillation counting. Binding results are

expressed as the radioligand binding inhibition or inhibition constant (Ki). Cellular functional assays were performed in the agonist mode using

CHO cells expressing human MT1 or MT2 receptors. The assay readout was cell impedance for MT1, measured by cellular dielectric spectroscopy,

and cAMP for MT2, measured by a ﬂuorometric method. Compounds were assessed in a log or semi-log dilution within the concentration range

0.01 nM to 100 μM. All compounds were full agonists. EC₅₀: half-maximal eﬀective concentration. NT: not tested. Experiments were performed

once in duplicate. Values are reported as the mean SEM.

explored in the alkylamide tail (Table 3). The increase of carbon

atoms from 2 to 3 in general reduced the binding aﬃnity for

MT1/MT2 (8a versus 14a, 8d versus 14b, 10a versus 15a and

10b versus 15b), but the functional potency was conserved in

some cases (MT1 and MT2 for 8a versus 14a and only MT2 for

10a versus 15a).

The chlorine atom was the only substituent investigated at the

5-position of the benzimidazole core. The addition of a chlorine

atom to that position was based on compound TIK-301 (N-

[(2R)-2-(6-chloro-5-methoxy-1H-indol-3-yl)propyl]-

acetamide). Besides adding lipophilicity to the benzimidazole

core, the chloro substitution blocks the formation of melatonin’s

main metabolite (6-hydroxymelatonin) and increases the half-

life of TIK-301.⁴¹The chloro group was well tolerated at MT1

and MT2 when analyzing the potency data for the

benzimidazole pairs with 2 carbon atoms in the alkylamide tail

(8d versus 10b and 8a versus 10a, Table 3). In contrast, the

insertion of a chlorine atom in the derivatives with 3 carbon

atoms in the alkylamide tail resulted in a lower activity at MT1

(14b versus 15b and 14a versus 15a, Table 3).

The small functional preference seen for compound 15a (2-

ethoxybenzimidazole derivative with a chloro group at the 5-

position and 3 carbon atoms in the alkylamide tail) toward the

MT2 receptor prompted us to analyze its binding mode at MT1

and MT2. The binding mode for compound 10b (2-

methoxybenzimidazole derivative with a chloro group at

position 5 and 2 carbon atoms in the alkylamide tail), which

displays a small opposite preference toward the MT1 receptor,

was also analyzed for comparison. Docking studies (Figure 1)

were conducted using the X-ray co-crystal structures with the

best available resolution: MT1 bound to ramelteon (PDB ID:

6ME62) and MT2 bound to 2-phenylmelatonin (PDB ID:

6ME6); these were the same structures previously used in

docking simulations with MT1/MT2 ligands.^34,35The basic

interaction requirements for MT1/MT2 binding are the

aromatic stacking with Phe178/Phe192 and the hydrogen

bonds with Asn162/Asn175 and Gln181/Gln194. The Tyr187/

Tyr200 residues that reside at the channel entrance of the

orthosteric binding site in MT1/MT2, respectively, adopt

diﬀerent conformations. The Tyr200 in MT2 makes a H-bond

interaction with Asn175 that constricts the channel entrance of

the orthosteric ligand-binding site, whereas the corresponding

Tyr187 in MT1 points to the lipid membrane.³⁵The

computational binding modes of compounds 10b and 15a

were very similar to the experimental binding modes of

ramelteon and 2-phenylmelatonin. However, 10b and 15a

displayed similar Glide docking scores at MT1 (−8.540 and

−9.307 kcal/mol, respectively) and MT2 (−9.306 and −9.384

kcal/mol), which were not able to explain the small preferences

of 10b for MT1 and 15a for MT2. The 2-ethoxy group in the

benzimidazole core of 15a lies in the same position as the 2-

phenylmelatonin phenyl ring in MT2. The binding pocket of

MT2 is larger than that of MT1 (approximately 50 Å³), with

most of the volume diﬀerences around the alkylamide tail and

the 2-position group region.³⁵The larger hydrophobic

subpocket in MT2 around the alkylamide tail and 2-ethoxy

group may explain the small preference of the bulkier compound

15a for MT2. In contrast, compound 10b has smaller structural

features that may ﬁt better in the smaller binding pocket of MT1,

such as the alkylamide tail with two carbons in the linker and the

2-methoxy group. It is also possible that the longer alkylamide

tail of 15a with three methylene groups may preclude optimal H-

bond interactions with MT1. This is in line with the gain of MT2

selectivity for ligands with longer ethyl and n-propyl amide

substitutions in the amide group,³⁰which could hinder the

formation of optimal H-bond interactions with the Gln181

residue at MT1.

Novel Benzimidazole Derivatives with Potent MT1

and MT2 Agonism Have Low Lipophilicity, Oral

Bioavailability, Peripheral Distribution, Good Selectiv-

1909

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 1. Binding poses of 10b and 15a in the human MT1 (PDB: 6ME2) and MT2 (PDB: 6ME6) receptors, obtained from docking studies using the

standard precision mode in Glide. The top-ranked poses for 15a (turquoise) according to Glide are represented in the MT1 (A) and MT2 (C)

receptors. The top-ranked poses for 10b (pink) according to Glide are represented in the MT1 (B) and MT2 (D) receptors. Hydrogen-bond

interactions are displayed as yellow dashed lines between the ligands and speciﬁc residues (Asn162 and Gln181 for MT1 and Asn175 and Gln194 for

MT2). The benzimidazole core also interacts via pi-interactions with phenylalanine residues in the binding site (Phe178 in MT1 and Phe192 in MT2).

ity, and Lack of CYP Inhibition. Physicochemical and in vitro

ADME experiments were performed for selected benzimidazole

derivatives and reference compounds to further characterize

their drug-like properties. In addition, in vivo PK studies were

conducted in rats to determine the plasma exposure of some

benzimidazole derivatives following oral and intravenous

administration as well as their brain-to-plasma distribution

ratios.

The logD values were experimentally obtained for melatonin,

ramelteon, agomelatine, and few benzimidazole derivatives; pK_a

values were also experimentally determined for most of the

latter. The conjugated acid of compound 4the direct

melatonin analoguehas a pK_avalue of 5.02, being mildly

acidic. Despite mainly existing in its neutral form at physiological

pH, the change from an indole to a benzimidazole ring still

reduces lipophilicity. Compound 4 is 0.5 log unit less lipophilic

than melatonin, i.e., 0.71 versus 1.19. The potent, non-selective

MT1/MT2 agonists with high brain exposure, ramelteon and

agomelatine, are indeed more lipophilic, displaying logD values

of 2.60 and 2.59. The addition of ethoxy or methoxy substituents

at the 2-position decreases the pK_avalues of the conjugated acids

of compounds 8a and 8d to 3.71 and 3.44, respectively, i.e., 1−

1.5 log units, lower than compound 4 (Table 4), in agreement

with the projected increase in acidity made by the DFT

calculations. The addition of ethoxy or methoxy substituents at

the 2-position also slightly increases lipophilicity, bringing the

logD values of compounds 8d and 8a to 0.94 and 1.26,

respectively (Table 4). The tricyclic dihydrofuran benzimida-

zole derivative 19a, substituted with an methoxy group at the 2-

position, has a pK_avalue of 3.46, similar to compounds 8a and

8d, and a logD value of 0.94, similar to the non-potent

benzimidazole analog (4). The addition of the electron-

withdrawing chloro substituent at the 5-position in compounds

15a and 10b further increases acidity, bringing the pK_avalues to

3.01 and 2.69, and also increases lipophilicity, bringing the logD

values to 1.99 and 1.44, respectively, but still ∼0.5−1.0 log unit

less lipophilic than ramelteon and agomelatine. The pK_avalues

of the conjugated acids of the benzimidazole derivatives 4, 8a,

8d, 15a, 10b, and 19a range from 2.69 to 5.02, indicating they

mostly exist in their neutral forms at physiological pH. In

addition, their logD values range from 0.71 to 1.99, slightly less

or more lipophilic than melatonin’s (logD = 1.19). In other

words, all benzimidazole derivatives mimic well melatonin’s

physicochemical properties and should have no problem to

diﬀuse across lipophilic membranes to bind to the MT1/MT2

receptors. Therefore, their activities could be ascribed as a

function of the speciﬁc interactions they form with each

receptor; the 2-substitution seems to be essential for that. The

1910

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Table 4. In Vitro Physicochemical and ADME Properties of Compounds 8a, 8d, 10a, 10b, 14a, 14b, 15a, 19a, and 19b

Metabolic stability in hepatocytes t_1/2(min) | CL_int

(μL/min/10⁻⁶cells)

Caco-2 P_app(× 10⁻⁶cm/s)

Compound

LogD

pK_a

A-B

B-A

K_S(μM)

>200

Human

Rat

Mouse

8a

1.26 0.09

3.71 0.02

23.8 12.9

41.3 7.8

90.0 30.0 |

12.3 4.1

>120 |

<8.2

101.5 18.5 |

10.1 1.9

>120 |

<8.2

>120 |

<8.2

>120 |

<8.2

>120 |

<8.2

>120 |

<8.2

>120 |

<8.2

32.8 19.2 |

46.0 27.0

65.6 10.4 |

15.5 2.5

34.2 17.8 |

39.6 20.7

114.7 6.7 |

8.7 0.5

36.0 27.0 |

63.1 47.5

75.5 44.5 |

20.0 11.8

66.3 53.7 |

43.5 35.3

83.3 36.7 |

14.7 6.5

73.9 46.1 |

21.9 13.7

91.5 28.5 |

11.9 3.7

70.6 49.4 |

27.4 19.2

68.0 38.2 |

27.9 18.6

74.9 45.1 |

20.7 12.5

8d

0.94 0.08

1.80 0.01

1.44 0.02

1.47 0.04

1.17 0.05

1.99 0.04

0.94 0.13

1.27 0.05

3.44 0.01

NT

21.3 9.6

38.0 20.2

17.9 3.5

22.5 9.6

18.8 8.6

23.3 10.7

12.8 4.9

15.8 5.6

25.8 5.6

30.0 2.2

49.5 22.3

26.7 2.8

38.1 17.1

44.8 13.0

30.2 6.9

39.9 25.6

>200

10a

10b

14a

14b

15a

19a

19b

197.0 3.0

198.4 1.7

>200

2.69 0.05

NT

43.5 9.5 |

23.8 5.2

>120 |

NT

198.6 1.4

181.4 0.6

>200

<8.2

3.01 0.01

3.46 0.01

NT

>120 |

<8.2

103.7 16.3 |

12.0 3.8

98.1 3.1 |

10.1 0.4

197.9 0.1

a

A-B and B-A P_app: apical to basolateral and basolateral to apical apparent permeabilities. K_S: kinetic solubility. Metabolic stability half-life (t_1/2) and

intrinsic clearance (CL_int). NT: not tested. Data are reported as mean SEM of two independent experiments in duplicate.

benzimidazole core oﬀers therefore a good opportunity to

achieve peripheral exposure due to its low lipophilicity, while

retaining potency at the MT1/MT2 receptors.

Table 5. In Vitro Pharmacology at MT1 and MT2 of

Compounds 8a, 8d, 10a, 10b, 14a, 14b, 15a, 19a, and 19b

a

Compound

MT1 agonism EC₅₀(nM)

MT2 agonism EC₅₀(nM)

We have then characterized the ADME properties of 11

analogs to obtain an initial assessment of their potential in vivo

PK and drug−drug interaction (DDI) proﬁles. We ﬁrst tested

CYP inhibition to mitigate potential CYP-mediated DDI risk

(Table S2). Out of the tested compounds, only 8b and 8c had

some inhibitory eﬀect at CYP2C19 (55% and 82% at 10 μM,

respectively) and were withdrawn from further investigation.

Table 4 shows the ADME properties for the remaining analogs

(8a, 8d, 10a, 10b, 14a, 14b, 15a, 19a, and 19b). All compounds

were permeable (>10 × 10⁻⁶cm/s) in Caco-2 cells and soluble

in PBS at pH = 7.4 (values close to 200 μM or higher). The

metabolic stability of selected compounds from our series (8a,

8d, 14a, and 10a) was initially assessed after incubation with

human liver microsomes (HLM) (Table S3) and compared with

HLM data for melatonin from the literature.⁴²The tested

benzimidazole derivatives were highly stable in HLM, showing

estimated half-lives of at least 320 min. They also showed

microsomal intrinsic clearance values (CL_int) of 21.2, 9.9, 9.6,

and 11.6 μL/min/mg, respectively, most of the values lower

than the reference value for melatonin (CL_int= 18.9 μL/min/

mg). Given the high stability of the compounds in a CYP-

predominant system (HLM), we decided to assess their

metabolic stability in a more complex in vitro metabolism

model by incubating the compounds with hepatocytes from

diﬀerent species. They had overall high metabolic stability in

human, rat, mouse, and dog hepatocytes (Table 4 and Table S2).

The nine key analogs were then retested in vitro to conﬁrm

their MT1/MT2 agonism and possible selectivity in two

independent cellular functional assay experiments. To allow a

direct head-to-head comparison with the MT2 assay, we have

also evaluated the MT1 potency of these analogs using cAMP

formation as the readout (Table 5). The MT2 data reported in

Table 5 is the result of two independent determinations with the

same ﬂuorometric protocol previously employed. These assays

8a

8d

0.040 0.002

0.042 0.004

0.035 0.002

0.063 0.008

0.036 0.002

0.233 0.014

0.937 0.204

0.042 0.001

0.037 0.004

0.106 0.073

0.210 0.146

0.101 0.075

0.321 0.277

0.185 0.096

0.887 0.541

0.141 0.080

0.049 0.011

0.035 0.009

10a

10b

14a

14b

15a

19a

19b

a

Cellular functional assays were performed in the agonist mode using

CHO cells expressing human MT1 or MT2 receptors. The assay

readout was cAMP, measured by DiscoverX HitHunter assay (MT1)

or a ﬂuorometric method (MT2). Compounds were assessed in a log

or semi-log dilution within the concentration range 0.01 nM to 100

μM. All compounds were full agonists. EC₅₀: half-maximal eﬀective

concentration. Data are reported as mean SEM of two independent

experiments in duplicate.

have conﬁrmed the subnanomolar potency of the nine key

compounds in MT1/MT2 as well as the small functional

preference (ca. 5-fold) of 10b and 15a toward MT1 and MT2,

respectively.

Table 6 displays the in vivo PK parameters of the nine key

compounds in mice and rats, after intravenous (1 mg/kg) and

oral dosing (10 mg/kg). We have also included reference in vivo

PK parameters for melatonin determined in previous studies in

rats^43,44for comparison. The series demonstrates decent oral

exposure in rodents. Saturation was observed for some

compounds, which displayed oral bioavailability greater than

100%. Three compounds, 15a, 10b, and 19a, were selected for

further characterization based on their good PK, high potency,

and diﬀerent functional proﬁles. Compared to melatonin,

compounds 15a, 10b, and 19a exhibited increased bioavail-

ability, half-life, and AUC following oral administration in rats.

1911

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

b

Table 6. In Vivo PK of Compounds 8a, 8d, 10a, 10b, 14a, 14b, 15a, 19a, and 19b in Rats and Mice

Intravenous

Oral

Species

(Strain)

Compound

CL

V_ss

t_1/2

C₀

F

t_1/2

C_max

AUC_0‑inf

Rats

8a

8d

57.8 2.7

65.4 5.0

53.3 2.1

29.8 1.0

48.1 2.6

26.0 3.8

16.6 0.8

34.6 1.3

35.2

0.69 0.04

0.92 0.05

0.43 0.03

0.52 0.06

0.58 0.01

0.65 0.06

0.32 0.02

0.96 0.02

1.05

0.21 0.02

0.26 0.02

0.14 0.01

0.29 0.02

0.25 0.01

0.52 0.01

0.41 0.1

0.71 0.09

0.33

1647 43

1578 164

2726 382

2167 182

2179 210

1544 192

3272 53

1492 93

934.6

15.7 6.1

13.2 0.9

112.3 19.9

72.3 3.4

15.9 1.0

36.5 1.5

108.0 0.1

88.2 11.0

53.5

0.50 0.16

0.61 0.14

0.67 0.06

1.05 0.35

0.94 0.38

1.14 0.31

1.06 0.31

1.56 0.12

0.53

519 97

337 70

527 176

343 20

3520 615

4053 190

555 37

2454 92

10,977 49

4303 535

2490

(Wistar

Han)

10a

10b

14a

14b

15a

19a

3535 74

3115 166

785 139

1610 21

6700 502

2300 424

2790

a

Rats (SD)

Melatonin

10b

15a

19a

38.7 1.2

15.2 1.8

33.7 4.0

70.7 2.7

38.2 0.2

29.4 2.1

25.5 0.8

31.2 0.2

18.8 0.7

10.3 0.1

46.2 0.9

77.5 2.4

0.56 0.01

0.27 0.02

1.29 0.13

0.51 0.01

0.77 0.01

0.43 0.01

0.33 0.03

0.49 0.01

0.45 0.05

0.22 0.01

1.18 0.06

1.45 0.33

0.24 0.01

0.26 0.01

0.61 0.03

0.12 0.01

0.29 0.02

0.18 0.02

0.19 0.03

0.24 0.02

0.30 0.01

0.28 0.01

0.31 0.01

0.29 0.04

1995

7

85.2 5.7

81.7 24.5

83.2 22.6

18.2 0.3

33.2 5.4

138 1.0

153 19.7

44.0 7.2

71.3 1.0

51.8 5.3

90.2 0.2

64.4 3.8

0.41 0.06

0.72 0.07

0.95 0.23

0.54 0.15

1.11 0.45

1.42 0.23

0.74 0.08

0.87 0.37

1.58 0.42

1.13 0.18

1.39 0.58

0.73 0.06

7955 155

9010 1490

3630 1100

836 123

1930 212

6490 728

8915 838

3405 105

5010 997

7550 99

3465 477

1575 102.5

3681 360

8743 1645

4031 637

4093 330

1217 73

1661 35

1314 68

2052 68

2659 245

1967 18

1956 209

4129 319

651 32

Mice (CD- 8a

429

7

1)

8d

10a

1449 236

7918 59

9999 1289

2356 387

6336 90.3

8336 854

10b

14a

14b

15a

19a

19b

3255

6

683 118

1388 82

a

The in vivo PK parameters for melatonin were obtained from the literature, estimated following intravenous (5 mg/kg) or oral (10 mg/kg)

administration in SD rats.^43,44Male CD-1 mice and Wistar Han and SD rats were dosed with compounds at 1 mg/kg (i.v.) and 10 mg/kg (p.o.).

Data are shown as mean SEM (n = 2 animals per group). CL: clearance (mL/min/kg). V_ss: apparent volume of distribution at the steady state

(L/kg). t_1/2: half-life (hours). F: oral bioavailability (%). C_max: maximum plasma concentration (ng/mL). C₀: concentration at time 0 (ng/mL).

AUC_0‑inf: area under the curve from time 0 extrapolated to inﬁnite time (h*ng/mL). NT: not tested. The PK parameters of melatonin were

extracted from the literature for comparison purposes.

b

Their brain-to-plasma (B/P) concentration ratios were

assessed in Wistar and SD rats (Table 7). The B/P ratio

experiments were initially conducted in Wistar rats using the

dose of 100 mg/kg p.o. to match the rat strain and maximum

dose that were used in the in vivo behavioral safety assessment.

These experiments were also performed in SD rats to match the

rat strain and doses (10 and 30 mg/kg) that were used in the in

vivo eﬃcacy experiments. The less lipophilic benzimidazole

compounds had improved PK (based on plasma exposure) and

substantially higher peripheral than central exposure (B/P ratios

approximately 0.1−0.2), when compared to ramelteon and

agomelatine (B/P ratios approximately 1 and 3, respectively),

conﬁrming our design hypothesis.

Rats Maintained on High-Fat Diet (HFD). Given the good

potency, selectivity and in vivo oral bioavailability of some

benzimidazole derivatives, we have decided to further explore

their in vivo eﬃcacy in a DIO rat model. The melatonin receptor

agonists 10b, 15a, and 19a were administered once daily to SD

rats (10 and 30 mg/kg p.o.) maintained on HFD, and their

eﬀects were compared with those of the SGLT2 inhibitor

dapagliﬂozin (5 mg/kg p.o.). Dapagliﬂozin was selected as a

comparator because our initial intent was to assess the eﬀects of

10b, 15a, and 19a on plasma glucose, insulin, and BW on a HFD

model in the context of type II diabetes mellitus (T2DM).

Dapagliﬂozin is an oral drug approved for glycemic control and

improvement of cardiovascular outcomes in T2DM patients.

This drug also normalizes blood glucose levels and reduces BW

in rat DIO models.^46,47Moreover, SGLT2 inhibitors are

currently being investigated as NAFLD/NASH therapies due

to the intricate relationship between glucose metabolism and

hepatic lipid accumulation/lipogenesis.⁴⁸Clinical trials have

already shown that SGLT2 inhibitors can reduce fatty liver

content and serum liver enzymes in patients with T2DM.⁴⁹

The selection of the oral doses for the MT1/MT2 agonists

was based on their unbound plasma concentration in SD rats

and their in vitro activity on the MT1/MT2 receptors (Figure

2); a coverage between 8 and 12 h above the in vitro functional

activities is expected at a dose of 30 mg/kg. Two vehicle-treated

control groups, fed either normal chow (Chow Vehicle) or HFD

(HFD Vehicle), were also included in the study. Animals were

maintained on chow or HFD during the acclimatization (days

−32 to 0) period to achieve a signiﬁcant HFD-induced BW gain

Compounds 10b, 15a, and 19a were further evaluated in the

commercially available selectivity panels SafetyScreen44 and

LeadProﬁlingScreen2 (Euroﬁns Pharma Discovery, Celle-

́

Levescault, France). SafetyScreen44 comprises a set of 44

targets involved in adverse eﬀects in the clinic,⁴⁵while the

LeadProﬁlingScreen2 consists of a panel of 80 GPCRs and ion

channels. At 10 μM, none of the compounds had binding aﬃnity

or enzyme inhibition above 50% for any of the targets, a standard

cutoﬀ for further investigation. Given their subnanomolar

activity at the MT1/MT2 receptors, these compounds are

expected to have at least a 10,000-fold selectivity over the tested

oﬀ-targets. This is important not only from a safety perspective

but also to eliminate potential confounding oﬀ-target eﬀects in

the in vivo eﬃcacy results.

Eﬀect of Dapagliﬂozin and Novel Benzimidazole

Derivatives with Potent Agonism at MT1 and MT2

Receptors on Body Weight (BW) and Food Intake (FI) of

1912

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

a

Table 7. Brain Exposure and B/P Ratio in Wistar and SD Rats of Compounds 10b, 15a, and 19a

Plasma concentration (μg/mL) | Brain concentration (μg/g) | B/P ratio

Animals (Dose)

Time (h)

0.37

Agomelatine

Ramelteon

10b

15a

19a

Wistar Han rats (100 mg/kg)

4.20 1.11 |

13.85 5.19 |

3.19 0.45

1.23 0.21 |

3.20 0.29 |

2.73 0.5

0.45 0.17 |

1.21 0.46 |

2.73 0.1

NT

3.16 0.71 |

2.85 0.56 |

0.91 0.02

3.45 0.55 |

3.03 0.65 |

0.86 0.08

1.15 0.08 |

1.19 0.19 |

1.02 0.1

NT

30.53 5.46 |

6.39 1.55 |

0.20 0.01

45.37 2.24 |

9.42 0.95 |

0.21 0.01

30.63 2.18 |

4.65 0.19 |

0.15 0.01

4.00 0.68 |

0.37 0.05 |

0.09 0.01

0.24 0.01 |

0.03 0.01 |

0.11 0.04

ND |

54.51 5.58 |

4.40 0.98 |

0.08 0.01

48.00 4.48 |

4.88 0.94 |

0.10 0.01

39.33 2.15 |

3.62 0.08 |

0.09 0.01

14.75 0.85 |

0.92 0.4 |

0.06 0.02

2.08 0.18 |

0.08 0.01 |

0.04 0.01

0.004 0.001 |

ND |

26.73 2.08 |

3.35 0.58 |

0.12 0.01

20.17 1.75 |

2.68 0.27 |

0.13 0.01

6.32 2.14 |

0.63 0.2 |

0.10 0.01

5.08 0.58 |

0.61 0.04 |

0.12 0.02

0.41 0.21 |

0.03 0.01 |

0.07 0.01

0.005 0.001 |

ND |

2

6

SD rats (10 mg/kg)

0.5

2

NT

8

ND |

ND

SD rats (30 mg/kg)

0.5

2

17.55 2.35 |

1.82 0.14 |

0.11 0.02

7.72 0.19 |

0.91 0.18 |

0.12 0.02

0.01 0.001 |

ND |

35.4 0.3 |

2.02 0.21 |

0.06 0.01

24.75 0.05 |

1.23 0.18 |

0.05 0.01

0.71 0.43 |

0.03 0.02 |

0.05 0.01

16.45 0.45 |

2.40 0.2 |

0.15 0.01

3.76 0.26 |

0.42 0.05 |

0.11 0.01

0.01 0.001 |

ND |

8

ND

a

Compounds were administered orally to male Wistar (100 mg/kg) and SD rats (10 and 30 mg/kg). The data are mean SEM of 2 (SD) or 3

(Wistar) animals per time point. NT: not tested. ND: not determined due to brain concentration below the lower limit of quantiﬁcation (5 ng/g).

Figure 2. Plasma concentration of 10b, 15a, and 19a in rats. Male SD rats (n = 2 per group) were orally treated with 10b, 15a, and 19a at the 10 and 30

mg/kg doses. Data are presented as mean SEM. Unbound plasma concentration after oral dosing was compared to the in vitro functional activity

(EC₅₀) at MT1 (A) and MT2 (B) receptors (cAMP formation as the readout).

before allocation to treatment groups (p < 0.001), following

study designs from the literature.^50,51

The daily BW data are presented in Figure 3. Animals from the

HFD Vehicle group steadily increased their BW from a starting

average of 459.9 g on day 1 to 612.4 g on day 57 (Figure 3A).

The rats from the Chow Vehicle group also gained weight

steadily, though less rapidly than the ones on HFD. Treatment

with compound 10b (10 and 30 mg/kg) signiﬁcantly reduced

BW gain on days 13 to 32 (p < 0.05) in comparison with the

HFD Vehicle control and dapagliﬂozin (5 mg/kg) on days 2 to

48, 50 to 54, and 56 (p < 0.05). Compared with the HFD

Vehicle, 10b signiﬁcantly reduced the weekly BW gain on weeks

1913

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 3. Eﬀect of 10b, 15a, and 19a on BW and FI of male SD rats maintained on HFD. Compounds 10b, 15a, 19a, dapagliﬂozin, and vehicle were

administered to rats fed normal chow or HFD. BW (A), weekly BW gain (B), average daily FI per measurement (C), and week (D) were analyzed. Data

are shown as adjusted means SEM of n = 9 animals (BW data, 10b at 10 mg/kg group), n = 10 animals (BW data, all other groups), or n = 5 cages (FI)

per group. The data were analyzed by one-way analysis of covariance (ANCOVA) with BW on day 1 (A, B) or average FI during the baseline period (C,

D) as the covariate, excluding the Chow Vehicle group. Treatment groups were compared with the HFD Vehicle by Williams’ tests (10b, 15a, and 19a)

or t tests (dapagliﬂozin), and statistically signiﬁcant diﬀerences are denoted by * (p < 0.05), ** (p < 0.01), and *** (p < 0.001). & (p = 0.051) and # (p

= 0.098) mean trends toward signiﬁcance in comparison with the HFD Vehicle. The 10b and dapagliﬂozin signiﬁcantly reduced BW and BW gain at

several time points. Dapagliﬂozin consistently increased FI, while isolated changes were observed for 10b, 15a, and 19a.

2 (10 and 30 mg/kg, p < 0.01), 3 (30 mg/kg only, p < 0.05), and

7 (10 and 30 mg/kg, p < 0.05), 15a at 30 mg/kg on week 2 (p <

0.05), and dapagliﬂozin on week 1 (p < 0.001) (Figure 3B). A

trend toward signiﬁcantly reduced BW gain from day 1 to day 57

versus HFD Vehicle was observed for 10b (10 and 30 mg/kg, p =

0.098) and dapagliﬂozin (p = 0.051).

assessed on days 23, 51, and 58/59 (Figure 4). Plasma glucose

(Figure 4A), insulin (Figure 4B), and HOMA-IR (Figure 4C)

were determined after a 4 h fast on days 23 and 51 and after a 16

h overnight fast on days 58/59 before an oral glucose tolerance

test (OGTT). Samples collected before the OGTT test (B1)

were used to analyze HOMA-IR and the liver enzymes alanine

(ALT) and aspartate (AST) aminotransferase (Figures 4D and

4E).

FI data are presented as kCal/rat/day consumed daily and as

average weekly intake (Figures 3C and 3D). The FI in the HFD

Vehicle group was broadly stable throughout the baseline and

dosing phase of the study at approximately 95−100 kCal/rat/

day. This value was ∼20% higher than that in the Chow Vehicle

animals. Apart from a few isolated ﬂuctuations, the overall

treatment with 10b, 15a, and 19a did not aﬀect daily FI.

Compound 10b signiﬁcantly decreased average daily FI only on

week 2 (10 and 30 mg/kg, p < 0.05), while a signiﬁcant increase

was observed for 15a (30 mg/kg) and 19a (30 mg/kg) on weeks

5 (p < 0.01 and p < 0.05, respectively) and 6 (p < 0.01, both

groups) in comparison with the HFD Vehicle. Dapagliﬂozin

signiﬁcantly increased daily FI from day 8 to 56 (p < 0.05) and

the average daily FI on weeks 2 to 7 (p < 0.01) compared with

the HFD Vehicle (Figures 3C and 3D, respectively). Average

daily food intake from day 1 to 56 was signiﬁcantly increased in

the animals treated with dapagliﬂozin and 19a (30 mg/kg) in

comparison with the HFD Vehicle control (p < 0.001 and p <

0.05, respectively).

Eﬀects of Dapagliﬂozin and Novel Benzimidazole

Derivatives with Potent Agonism at MT1 and MT2

Receptors on Glucose Tolerance and Liver Enzymes in

Rats Maintained on HFD. To further describe the metabolic

eﬀects of the benzimidazole derivatives and dapagliﬂozin, we

have assessed the levels of glucose, insulin, and liver enzymes in

HFD rats treated with these compounds. The eﬀects of 10b,

15a, 19a, and dapagliﬂozin on basal glucose and insulin were

On day 23, no signiﬁcant diﬀerences among the groups were

observed. On day 51, one-way ANOVA revealed signiﬁcant

diﬀerences for plasma glucose (F(8,80) = 4.24, p < 0.001),

insulin (F(8,80) = 5.93, p < 0.0001), and HOMA-IR (F(8,80) =

6.36, p < 0.0001). The HFD Vehicle animals had signiﬁcantly

increased plasma glucose (+17.3%, p < 0.05), insulin (+54.8%, p

< 0.05), and HOMA-IR (+78.4%, p < 0.01) compared with the

Chow Vehicle group. Post hoc comparisons with the HFD

Vehicle group also revealed that compound 10b signiﬁcantly

reduced plasma glucose at 10 mg/kg (−16.4%, p < 0.05) and 30

mg/kg (−16.9%, p < 0.01). However, only the 30 mg/kg dose

reduced HOMA-IR (−36.1%, p < 0.05). Compound 19a only

signiﬁcantly reduced plasma glucose at 10 mg/kg (−13.8%, p <

0.05), and no signiﬁcant diﬀerences were seen for compound

15a. Dapagliﬂozin signiﬁcantly decreased plasma glucose

(−21.5%, p < 0.001), plasma insulin (−56.7%, p < 0.0001),

and HOMA-IR (−66.8%, p < 0.0001) compared with the HFD

Vehicle. On days 58/59, signiﬁcant diﬀerences among the

groups were evidenced for plasma glucose (F(8,560) = 53.7, p <

0.001) and insulin (F(8,560) = 27.1, p < 0.0001) using a two-

way ANOVA and for HOMA-IR (F(8,80) = 5.22, p < 0.0001)

using a one-way ANOVA. The post hoc test showed a trend

toward a signiﬁcant increase in HOMA-IR for HFD Vehicle

animals compared with Chow Vehicle (+112.8%, p = 0.067).

Compounds 10b, 15a, or 19a did not exhibit signiﬁcant eﬀects

1914

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 4. Eﬀect of the novel melatonin receptor agonists 10b, 15a, and 19a on plasma glucose (A), insulin (B), HOMA-IR (C), and liver enzymes (D,

E) levels of male SD rats maintained on HFD. Compounds 10b, 15a, 19a, dapagliﬂozin, or vehicle were dosed to animals fed normal chow or HFD, and

the parameters were assessed on days 23, 51, and 58/59 (B1 sample from the OGTT). The plasma ALT and AST levels were only assessed on days 58/

59. Data are shown as adjusted means SEM, n = 9 for the 10b at the 10 mg/kg group and n = 10 for the remaining groups. The statistical analysis was

̌

́

done using one-way ANOVA followed by Holm−Sidak’s multiple comparisons test. Statistically signiﬁcant diﬀerences against the HFD Vehicle group

are denoted by * (p < 0.05), ** (p < 0.01), *** (p < 0.001), and **** (p < 0.0001). # (p = 0.067) means a trend toward signiﬁcance in comparison with

the HFD Vehicle. Dapagliﬂozin reduced plasma glucose, insulin, and HOMA-IR on days 51 and 58/59, and 10b reduced plasma glucose (both doses),

insulin (30 mg/kg), and HOMA-IR (30 mg/kg) on day 51 compared with the HFD Vehicle.

in comparison with the HFD Vehicle group. Conversely,

dapagliﬂozin signiﬁcantly reduced plasma glucose (−32.0%, p

< 0.01) and HOMA-IR (−77.7%, p < 0.01). Treatment with

10b, 15a, 19a, or dapagliﬂozin did not signiﬁcantly aﬀect the

AST or ALT levels when compared with the HFD Vehicle

(Figure 4D,E).

glucose administration (B2, which was 1 h after B1), and at 10,

20, 30, 60, and 120 min after glucose administration. Glucose

(Figure 5A) and insulin (Figure 5D) concentrations were

evaluated over time, and a two-way ANOVA test revealed

signiﬁcant diﬀerences among the groups for both parameters

(F(8,560) = 53.7, p < 0.0001 and F(8,560) = 27.1, p < 0.0001,

respectively). Signiﬁcant diﬀerences were also observed with

one-way ANOVA when analyzing glucose and insulin

The OGTT data are presented in Figure 5 and Figure S1.

Plasma was collected prior to treatment dosing (B1), prior to

1915

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 5. Eﬀect of the novel melatonin receptor agonists 10b, 15a, and 19a on glucose tolerance of male SD rats maintained on HFD. Compounds

10b, 15a, 19a, dapagliﬂozin, and vehicle were administered to animals fed normal chow or HFD. Data are shown as means SEM (n = 9 for the 10b at

the 10 mg/kg group and n = 10 for the remaining groups). The plasma glucose (A) and insulin (D) curves are shown for 120 min following the oral

glucose administration. The areas under the curve (AUC) for glucose (B, C) and insulin (E, F) are shown for the 0−60 or 0−120 min intervals,

̌

́

respectively. The statistical analysis was performed using two-way (A and D) or one-way (B, C, E, F) ANOVA followed by Holm−Sidak’s multiple

comparison tests. Statistically signiﬁcant diﬀerences against the HFD Vehicle group are denoted by * (p < 0.05), ** (p < 0.01), *** (p < 0.001), and

**** (p < 0.0001). Dapagliﬂozin treatment exhibited lower glucose and insulin concentrations at several time points, and 10b (30 mg/kg) showed

reduced plasma glucose at 30 min after the glucose load compared with the HFD Vehicle.

AUC_{0‑60 min}(F(8,80) = 12.3, p < 0.0001, Figure 5B, and F(8,80)

= 4.9, p < 0.0001, Figure 5E, respectively) and AUC_{0‑120 min}

(F(8,80) = 14.2, p < 0.0001, Figure 5C, and F(8,80) = 5.5, p <

0.0001, Figure 5F, respectively).

signiﬁcantly increased glucose AUC_{0‑60 min}and AUC_{0‑120 min}

(+16.5%, p < 0.05 and +15.2%, p < 0.05, Figure 6B,C) when

compared with the HFD Vehicle (Figures 5B and 5C). No

signiﬁcant eﬀect on glucose AUC_Bwas observed with any test

compound (Figure S1). Except for compound 15a (10 mg/kg)

at 20 min (+49.9%, p < 0.01), no signiﬁcant eﬀect of 10b, 15a, or

19a was observed on plasma insulin levels at baseline or during

the OGTT in comparison with the HFD Vehicle (Figure 5D).

Dapagliﬂozin signiﬁcantly reduced plasma glucose at the

baseline B1 (−32%, p < 0.01) and B2 samples (−45.3%, p <

0.0001) and at all time points from the OGTT (p < 0.01)

compared with the HFD Vehicle (Figure 5A). The dapagliﬂozin

group also exhibited reduced plasma insulin at B2 (−83.6%, p <

0.05) and throughout the OGTT test (p < 0.01, Figure 5D) in

comparison with the HFD Vehicle animals. The glucose and

insulin AUC_{0‑60 min}(−23.0%, p < 0.001 and −42.7%, p < 0.05,

respectively) and AUC_{0‑120 min}(−24.4%, p < 0.0001 and

−42.8%, p < 0.5, respectively) were signiﬁcantly reduced for

dapagliﬂozin-treated animals when compared with the HFD

Vehicle (Figure 5B,C,E,F).

The plasma glucose of the HFD Vehicle group was 6.42 mM

at B2 (prior to the administration of the glucose load) and rose

rapidly to a peak of 11.67 mM at 30 min before falling to 9.15

mM at 120 min (Figure 5A). A similar pattern was also observed

for plasma insulin, which was 1.32 ng/mL at B2, rose to 3.85 ng/

mL at 30 min, and returned to normal values at 120 min (Figure

5D). Plasma glucose levels were signiﬁcantly (p < 0.01) higher in

the HFD Vehicle group at 30, 60, and 120 min and insulin solely

at 30 min (p < 0.01) when compared with the Chow Vehicle.

The treatment with 10b, 15a, and 19a at 10 and 30 mg/kg had

no eﬀect on plasma glucose and insulin at B1 or B2 compared

with the HFD Vehicle (Figure 5A). During the OGTT, plasma

glucose was signiﬁcantly decreased by 10b (30 mg/kg) at 30 min

(−14.0%, p < 0.05). In contrast, when compared with the HFD

Vehicle group, there were signiﬁcant increases in glucose levels

in animals treated with compound 15a at 10 min (+31.0% at 30

mg/kg only, p < 0.001) and 20 min (+20.1% at 10 mg/kg, p <

0.01 and +18.4% at 30 mg/kg, p < 0.05), compound 10b at 10

min (+27.5% at 10 mg/kg, p < 0.01), and compound 19a (30

mg/kg) at 30 and 60 min (+14.9%, p < 0.05 and +25.1%, p <

0.001, respectively). Compound 19a at 30 mg/kg also

Eﬀect of Dapagliﬂozin and Novel Benzimidazole

Derivatives with Potent Agonism at MT1 and MT2

Receptors on Body Composition and Tissue Weights in

Rats Maintained on HFD. The weights of brown fat pad,

epididymal fat pad, liver, and gastrocnemius muscle were

1916

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Figure 6. Eﬀect of 10b or dapagliﬂozin on liver triglycerides (A, B), glycogen (C, D), and histopathology (E−H) of male SD rats maintained on HFD.

Compound 10b, dapagliﬂozin, or vehicle was dosed to animals fed chow or HFD, and parameters were assessed on days 65/66. Data are shown as

means SEM (A−D) or raw data including mean SEM (E−H), n = 9 for the 10b at the 10 mg/kg group and n = 10 for the remaining groups. The

statistical analysis of liver triglycerides (A, B), glycogen (C, D), and histopathology data (E−H) was performed by one-way ANOVA followed by

̌

́

Holm−Sidak’s multiple comparison tests. Signiﬁcant diﬀerences are denoted by * (p < 0.05), ** (p < 0.01), and *** (p < 0.001) and by † (p < 0.05),

†† (p < 0.01), and ††† (p < 0.001) for the comparisons against the HFD Vehicle or the Chow Vehicle groups, respectively. # p = 0.20 for the

comparison between the 10b (30 mg/kg) and the HFD Vehicle groups. Treatment with 10b reduced liver triglycerides and glycogen and improved

steatosis compared with HFD Vehicle. Dapagliﬂozin only reduced liver triglycerides concentration compared with HFD Vehicle.

assessed, and the data are presented in Table S4. The results

were analyzed in two ways: (i) BW on day 1 was used to correct

for diﬀerences in the BW at the start of the study (D1-corrected

analysis) and (ii) the terminal BW (days 65/66) was used to

correct for changes due to weight loss (D65/66-corrected).

diﬀerences were seen between the HFD Vehicle and Chow diet

groups. For the epididymal fat weight, the pairwise comparison

between these groups revealed a signiﬁcant increase in the D1

analysis (+50%, p < 0.05) and a trend toward a signiﬁcant

increase in the D65/66 analysis (+44.4%, p = 0.062). For the

liver weight, there was a trend of signiﬁcant increase in the HFD

Vehicle versus Chow diet in the D1 analysis (+15.7%, p = 0.066)

and a signiﬁcant increase in the D65/66 analysis (+13.4%, p <

0.05). Dapagliﬂozin treatment also led to a nearly signiﬁcant

increase in liver weight compared with the HFD Vehicle when

̌

́

One-way ANOVA followed by Holm−Sidak’s post hoc tests

showed diﬀerences between the groups for the epididymal fat

and liver weights on D1 (F(8,80) = 2.00, p = 0.0562 and F(8,80)

= 3.36, p < 0.01) and D65/66-corrected analyses (F(8,80) =

2.17, p < 0.05 and F(8,80) = 5.31, p < 0.0001). Most of the

1917

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

the data were corrected for the terminal BW (+10.8%, p =

0.053). Apart from these changes, no signiﬁcant eﬀect on the

weight of the selected tissues was observed for 10b, 15a, 19a, or

dapagliﬂozin compared with the HFD Vehicle, irrespective of

the correction.

0.001 and F(4,44) = 4.83, p < 0.01, respectively) and a diﬀerence

at the limit of statistical signiﬁcance (F(4,44) = 2.58, p = 0.050)

for lobular inﬂammation. Pathology scores for steatosis (1.4

versus 0, p < 0.0001), lobular inﬂammation (0.6 versus 0, p <

0.05), and total NAS (2.3 versus 0, p < 0.001) were signiﬁcantly

increased in the HFD Vehicle group in pairwise comparisons

versus the Chow Vehicle control, whereas hepatocellular

ballooning was not signiﬁcantly aﬀected. Remarkably, com-

pound 10b at 30 mg/kg signiﬁcantly reduced steatosis by 57% (p

< 0.05) when compared with the HFD Vehicle group (Figure

6F). A trend toward a signiﬁcant decrease in total NAS value was

observed between this group and the HFD Vehicle animals (p =

0.20, mean value of 1.3 versus 2.3). In contrast, dapagliﬂozin-

treated animals exhibited signiﬁcantly increased scores for total

NAS and steatosis when compared with the Chow Vehicle

control (1.5 versus 0, p < 0.05 and 0.9 versus 0, p < 0.05,

respectively), which were not statistically diﬀerent than the ones

observed for the HFD Vehicle group.

Metabolic Proﬁling and Early Safety Assessment of

Compound 10b. We have also investigated the metabolic

proﬁle and mutagenic potential of 10b as a step further in the

characterization of this lead compound. The metabolic proﬁling

of 10b after a 120 min incubation with human hepatocytes was

performed using HPLC-MS analysis. The summarized analytical

information about each metabolite is shown in Table S6, such as

HPLC retention time, relative peak area, calculated exact mass,

and experimental m/z data. Five potential metabolites were

identiﬁed (M1−M5), and the main proposed metabolic

pathways were O-demethylation followed or not by O-

glucuronidation (Figure S3). HPLC chromatograms are also

included in Figure S3. The proposed chemical structures of the

metabolites and the MS spectra are shown in Figure S4. Based

on UV peak areas, 71.9% of the parent compound remained after

incubation, and the main metabolite was formed via O-

demethylation (14.8% of the total UV area).

Body composition was determined by FoodScan analysis,

which measures water, protein, and fat content and the ﬁnal

carcass weight (Table S5). One-way ANOVA followed by

̌

́

Holm−Sidak’s post hoc test revealed a trend of signiﬁcant

diﬀerence between the groups just for carcass weight (F(8,80) =

1.98, p = 0.059), and the pairwise comparison disclosed

signiﬁcant diﬀerences for this parameter only between the

HFD Vehicle versus Chow Diet groups (+17.0%, p < 0.05). No

eﬀects of dapagliﬂozin, 10b, 15a, or 19a were observed in ﬁnal

carcass weight, water, protein, or fat content compared with the

HFD Vehicle group.

Eﬀects of Compound 10b and Dapagliﬂozin on Liver

Triglyceride Levels, Glycogen Content in Muscle and

Liver, and Hepatic Steatosis in Rats Maintained on HFD.

As only 10b, but not 15a or 19a, displayed some eﬀect on BW

gain, plasma glucose levels, and HOMA-IR in the HFD rat study,

we further characterized the liver and muscle eﬀects of this

compound and compared with the Chow Vehicle, HFD Vehicle,

and dapagliﬂozin groups. The parameters evaluated included

liver triglyceride levels, glycogen content of both liver and

muscle, and the histopathological analysis of the liver (Figure 6).

̌

́

One-way ANOVA followed by Holm−Sidak’s post hoc test

revealed statistically signiﬁcant between-group diﬀerences for

liver triglycerides (F(4,44 = 7.34, p < 0.0001 for concentration

and F(4,44) = 6.94, p < 0.001 for total content analysis, Figures

6A and 6B) and liver glycogen (F(4,44 =,6.84, p < 0.0001 for

percentage and F(4,44) = 5.72, p < 0.001 for total content,

Figures 6C and 6D), but not for muscle glycogen. In pairwise

comparisons, liver triglyceride concentration and total content

were markedly increased (+473%, p < 0.0001 and +409%, p <

0.0001, respectively) in the HFD Vehicle animals when

compared with the Chow Vehicle control. Interestingly,

treatment with 10b at 10 and 30 mg/kg signiﬁcantly reduced

the liver triglyceride total content (−41.3%, p < 0.05 and

−55.1%, p < 0.01, respectively) and concentration (−34.4%, p <

0.05 and −51.6%, p < 0.01, respectively) when compared with

the HFD Vehicle. In contrast, dapagliﬂozin treatment

signiﬁcantly reduced only the hepatic triglyceride concentration

when compared with the HFD Vehicle control (−35.4%, p <

0.05).

Compound 10b did not induce mutagenicity in the Ames test

in the presence or absence of a metabolic activation system (rat

liver S9), up to a concentration of 100 μM. In the functional

observational battery (FOB), 10b did not cause any detectable

behavioral changes over the dose range of 25−100 mg/kg p.o.

(Figure S5). This test evaluates more than 30 parameters from

the autonomic, neuromuscular, sensorimotor, and behavioral

domains and is useful to assess safety liabilities in both non-

central nervous system (CNS)- and CNS-targeted com-

pounds.^52,53

The glycogen content and concentration in the liver (Figures

6C and 6D) and muscle (Figure S2) were similar between the

Chow Vehicle and HFD Vehicle groups. Compound 10b at 10

and 30 mg/kg signiﬁcantly reduced the liver glycogen content

(−36.6% and −36.6%, respectively, p < 0.05) and concentration

(−42.8% and −40.8%, respectively, p < 0.05) when compared

with the HFD Vehicle (Figures 6C and 6D). No eﬀect of either

10b or dapagliﬂozin was observed on muscle glycogen content

or concentration in comparison with the HFD Vehicle control.

Samples of the animal livers were also sectioned, stained, and

scored for steatosis, lobular inﬂammation, and hepatocellular

ballooning, and the overall non-alcoholic fatty liver disease

(NAFLD) activity score (NAS) was determined (Figure 6E−

H). The liver of the control animals fed normal chow had no

pathology scores (zero) for steatosis, lobular inﬂammation, or

hepatocellular ballooning and no score for total NAS. The one-

way ANOVA revealed a signiﬁcant diﬀerence between the

groups for liver steatosis and total NAS (F(4,44) = 6.73, p <

DISCUSSION AND CONCLUSIONS

■

We have reported here novel benzimidazole derivatives bearing

potent activity at MT1 and MT2 receptors and good oral

bioavailability, with substantially higher peripheral than brain

exposure. In addition, we have also evaluated the metabolic

eﬀects of compounds 10b, 15a, and 19a in the HFD rat model,

following chronic oral administration.

The 2-alkoxybenzimidazoles were designed to exhibit

improved metabolic stability and increased peripheral exposure

when compared to melatonin and marketed melatonin receptor

agonists. To accomplish that, the design strategy included (i)

lowering lipophilicity by replacing the indole with the

benzimidazole core, which is more polar, and (ii) blocking

oxidation of the benzimidazole 5-position, which is equivalent to

the 6-position of melatonin, its main soft spot. Exogenous

melatonin has a short plasmatic half-life and low and highly

variable bioavailability after oral administration in humans,

1918

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

limiting its applicability as a drug therapy.⁵⁴The lead

compounds 10b, 15a, and 19a had enhanced oral bioavailability,

half-life, and AUC values in rats when compared to melatonin

and displayed overall good metabolic stability in hepatocytes

across species, projecting an improved PK proﬁle in humans.

A 2 month oral treatment with 10b led to a signiﬁcant

decrease in BW gain on days 13 to 32, on weeks 2, 3, and 7 (days

8 to 22 and 43 to 50) and a trend toward a signiﬁcant decrease in

BW gain between days 1 to 57 in comparison with the HFD

vehicle group. This compound also caused marked reductions in

liver glycogen, triglyceride levels, and steatosis in the HFD rat

model as well as improved glucose plasma and HOMA-IR on

day 51. In contrast, dapagliﬂozin did not signiﬁcantly alter liver

steatosis or liver triglyceride total content and glycogen contents

but signiﬁcantly reduced BW gain and improved glycemic

control, insulin sensitivity, and liver triglyceride concentration.

Dapagliﬂozin only produced a slightly more signiﬁcant

reduction of BW gain than 10b at 10 and 30 mg/kg on day 57

(4.3%, 4.1%, and 3.5%, respectively), even though caloric intake

was increased by 13.8% overall for dapagliﬂozin-treated animals

(days 1 to 56). This result is consistent with the ﬁnding that the

weight loss eﬀect of dapagliﬂozin in DIO rats is somewhat oﬀset

by the compensatory hyperphagia induced by urinary glucose

excretion.⁴⁸In patients with T2DM, chronic treatment with an

SGLT2 inhibitor increases caloric intake by 10−15%, which in

turn compensates 9 out of the 11 kg in the expected glycosuria-

induced BW loss over 2 years.⁵⁵

Despite the treatments with 10b and dapagliﬂozin that

resulted in similar reductions in BW gain, only the former

reduced both liver triglyceride content and concentration and

improved the degree of hepatic steatosis. Conversely, dapagli-

ﬂozin treatment led to a nearly signiﬁcant unwanted increase in

liver weight when compared with the HFD Vehicle and the data

were corrected for the terminal BW (+10.8%, p = 0.053). These

ﬁndings could be interpreted as a selective and direct action of

10b on the liver in contrast to a non-speciﬁc eﬀect resulting from

the overall reduction in BW gain.

Melatonin supplementation also decreases BW gain, hepatic

steatosis and inﬂammation, and liver triglycerides in rodent

NAFLD/NASH models^56,57due to its action on MT1/MT2

receptors and antioxidant activity.⁵⁸Moreover, a pilot clinical

study has shown beneﬁcial eﬀects after a 12 week treatment with

melatonin on liver enzymes in patients with NASH.⁵⁹These

preliminary results encourage the execution of randomized

placebo-controlled clinical trials in NAFLD/NASH patients

treated with melatonin receptor agonists with improved PK and

preferred peripheral exposure, over a longer period and having

liver histology as the primary endpoint.

MT1/MT2 agonist with substantially greater peripheral than

central exposure and robust modulation of hepatic lipid

metabolism in the DIO rat model, oﬀers an important

contribution to the ﬁeld. The lack of any behavioral changes

for compound 10b up to the dose of 100 mg/kg in the FOB test

in rats further substantiates its CNS safety.

The eﬀects of 10b on liver metabolism suggest that the

activation of melatonin receptors in peripheral tissues could

oﬀer an interesting therapeutic strategy for metabolic diseases.

Studies in genetically ablated mice for MT1 (MT1^−/−) or MT2

(MT2^−/−) fed with HFD showed diﬀerent eﬀects of these

receptors on glucose and insulin metabolism. MT1^−/−mice

exhibit a robust metabolic phenotype, with higher cumulative

weight gain, hyperglycemia, and marked insulin resistance.⁶²In

contrast, MT2^−/−mice displayed reduced hepatic insulin

sensitivity but increased insulin release.^16,63Moreover, the

regulation of MT1 and MT2 has important diﬀerences upon

melatonin stimulation. The activation of MT2 desensitizes

cAMP response, whereas MT1 leads to cAMP supersensitiza-

tion in the subsequent withdrawal period.^64,65The nocturnal

activation of MT1 in mice also modulates hepatic insulin

sensitivity during the day, whereas decreased MT1 expression in

human liver is linked to poorly controlled diabetes.⁶⁵It is then

possible that the positive eﬀects of compound 10b in the liver

could be linked to its small functional preference toward MT1

(Table 4).

Obesity contributes to the development of NAFLD and

eventually NASH, characterized by macrovesicular steatosis,

hepatocyte ballooning, and ﬁbrosis. Therefore, we postulate that

compound 10b might add value to the treatment of NAFLD/

NASH. Since animals in the 3 month HFD group exhibited

marked steatosis but only mild liver ﬁbrosis, additional studies

should be conducted in models with established hepatic ﬁbrosis

to conﬁrm the potential of 10b in the treatment of liver

metabolic diseases.

In summary, we described here a novel potent and

peripherally preferred melatonin receptor agonist, compound

10b, that manages weight gain and signiﬁcantly reduces liver

triglyceride levels and histology-determined steatosis in the

HFD rat model. Compound 10b was also devoid of CYP

inhibition, hERG binding, and genotoxicity and exhibited an

extremely clean selectivity proﬁle, low CNS exposure, and no

behavioral concerns at doses up to 100 mg/kg p.o. Based on its

eﬃcacy, oral pharmacokinetics, and safety proﬁle, compound

10b was nominated as ACH-000143 and selected for further

investigation in the path toward a drug candidate against

NAFLD/NASH.

Although the short-term use of melatonin supplements is

considered safe, its long-term safety still requires further

investigation.⁶⁰The most common side eﬀects of melatonin in

humans are related to its action in the CNS, such as headache,

sleepiness, drowsiness, and irritability. Indeed, the cerebrospinal

ﬂuid/plasma concentration ratio of exogenous melatonin

administered to rats is reported to be 0.38²⁷and 0.46,⁶¹but

brain levels can reach up to 3.8-fold higher values than in the

cerebrospinal ﬂuid, which is in line with the range of dose-related

behavioral eﬀects of melatonin in rodents.²⁷Such CNS-related

side eﬀects could be mitigated by reducing the brain exposure of

MT1/MT2 agonists when targeting peripheral diseases.

However, few to none studies have described less lipophilic

melatonergic ligands targeting preferentially peripheral MT1/

MT2 receptors. Therefore, the discovery of 10b, a potent full

EXPERIMENTAL SECTION

■

Drugs and Formulation. Agomelatine, ramelteon, melatonin, and

dapagliﬂozin were purchased from Sigma-Aldrich (St. Louis, MO, USA

and Dorset, UK). Compounds were formulated in 5% dimethyl

sulfoxide (DMSO), 30% polyethylene glycol 400 (PEG400), and 65%

distilled water for the PK studies in CD-1 mice and Wistar Han rats. If

the p.o. formulation was not a clear solution, 0.2% hydroxypropyl

methylcellulose (HMPC) was added to increase suspension homoge-

neity. In the HFD study, compounds were formulated in the vehicle for

the control groups, which was 1% DMSO, 1% Tween 80, and 98%

distilled water with 0.2% HPMC. This formulation was also used for the

oral PK studies in Sprague Dawley (SD) rats. For the intravenous (i.v.)

administration in SD rats, compounds were formulated in 5% DMSO,

30% PEG400, and 65% distilled water. Compound 10b was formulated

in distilled water with 0.2% HPMC for the FOB study.

1919

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Animal Procedures. All experiments were conducted in

accordance with institutional guidelines as deﬁned by the Institutional

Animal Care and Use Committee for U.S. institutions and the

recommendations of the Association for Assessment and Accreditation

of Laboratory Animal Care International (AAALAC). Male CD-1 mice

(25−35 g) from Lingchang (Shanghai, China) and male Wistar Han

rats (weight range of 230−270 g) and SD rats (7−9 weeks of age) from

Vital River (Beijing, China) were used in the PK studies. Male Wistar

Han (Elevage Janvier, Le Genest-Saint-Isle, France) and male SD rats

(Charles River, Margate, UK), with a weight range between 200 and

250 g, were employed in the FOB and the HFD eﬃcacy studies,

respectively. Except if stated otherwise in the study protocol, animals

were housed in pairs at a temperature of 21 4 °C, relative humidity

typically at 55 20%, artiﬁcial lighting between 7:00 and 19:00, and

food and water available ad libitum. Safety and eﬃcacy in vivo

experiments were performed at Porsolt (Le Genest-Saint-Isle, France)

and RenaSci Ltd. (Nottingham, UK), respectively. Experiments were

performed under authorizations from Porsolt’s (2016082313175580)

and RenaSci’s (PPL30/3208) ethics committees.

In Vitro Pharmacology. The novel benzimidazole derivatives were

assessed at the MT1 and MT2 receptors following radioligand binding

and cellular functional assays developed elsewhere.⁶⁶⁻⁶⁸All com-

pounds were initially screened at 1 μM in the binding assays, once in

duplicate, and a percentage of binding inhibition equal or greater than

95% for one of the targets was used as a selection criterion for further

testing. Then, independent binding and cellular functional assays were

performed, once in duplicate, in a dose−response fashion for the

selected compounds in the single-point assay to determine the

concentrations producing half-maximal inhibition of control speciﬁc

binding (IC₅₀) and half-maximal agonism response (EC₅₀) at MT1 and

MT2. Nine key compounds were retested in the in vitro functional

assays in a dose−response fashion to conﬁrm their MT1/MT2 agonism

and possible selectivity in two independent experiments. The

compounds selected for the in vivo eﬃcacy study were tested in the

selectivity panels SafetyScreen44 and LeadProﬁlingScreen2 (Euroﬁns

Pharma Discovery, France). A total of 93 targets, including enzymes,

ions channels, transporters, and receptors, were tested at 10 μM, once in

duplicate.

The radioligand binding assays were conducted using CHO cells

expressing human recombinant MT1 or MT2 receptors.^66,68

Compounds were incubated in a log or semi-log dilution within the

concentration range 0.01 nM to 100 μM with the cells and the

radioligand ([¹²⁵I]2-iodomelatonin) at 0.01 nM for MT1 and 0.05 nM

for MT2. Samples were incubated for 60 min at room temperature for

the MT1 binding assay and 120 min at 37 °C for the MT2 binding

assay. Nonspeciﬁc binding was evaluated through the incubation with

melatonin at 1 μM for MT1 and MT2. After incubation, the % of

radioligand binding inhibition was determined via a scintillation

counting method. The radioligand binding experiments were

performed once in duplicate. The IC₅₀was determined by a nonlinear

regression analysis of the concentration−response curves ﬁtted to a Hill

equation curve. The inhibition constants (K_i) were calculated using the

Cheng−Prusoﬀ equation.

The cellular functional assay was performed in the agonist mode

using CHO cells recombinantly expressing human MT1 or MT2

receptors. The MT1 agonism was initially assessed using a cell

impedance assay,⁶⁷measured by cellular dielectric spectroscopy, and,

later, conﬁrmed with the DiscoverX HitHunter cAMP XS+ assay

(Fremont, CA, USA), which quantiﬁes cAMP levels. The MT2 receptor

agonism assay readout was cAMP, measured by a ﬂuorometric

method.⁶⁶Cells were incubated with the compounds at 28 °C (MT1

impedance assay) or 37 °C (MT1/MT2 cAMP assays). In the

DiscoverX HitHunter assay, the incubation was conducted in the

presence of forskolin (20 μM), an adenylyl cyclase activator. cAMP-β-

galactosidase fragment, anti-cAMP antibody, β-galactosidase enzyme

complement, and β-galactosidase substrate were added to the assay

medium, and the subsequent reaction occurred at 37 ° C. The

chemiluminescent signal from the hydrolyzed substrate was quantiﬁed.

Compounds were assessed in a log or semi-log dilution within the

concentration range 0.01 nM to 100 μM. The results were obtained as a

percent of melatonin maximum response. The MT1 cell impedance

assay was performed once in duplicate. Two independent experiments

in duplicate were conducted for the MT1 and MT2 cAMP assays. The

EC₅₀values were determined by the nonlinear regression analysis of the

concentration−response curves using the Hill equation curve ﬁtting.

The in vitro pharmacology experiments were conducted at Euroﬁns

́

Cerep (Celle-Levescault, France) and Euroﬁns DiscoverX (Fremont,

USA).

Computational Studies. pK_aCalculations. The computational

̈

studies were performed within the Schrodinger suite (Maestro,

̈

Schrodinger, LLC, New York, NY, 2019). The pK_avalues were

calculated using Jaguar³⁷(Jaguar, Schrodinger, LLC, New York, NY,

2020) software. The ligands were built using CANVAS (Canvas,

̈

Schrodinger, LLC, New York, NY, 2019) and exported to MAESTRO

r.2019-1. The Ligand Preparation⁶⁹tool (LigPrep) was used to

generate energy-minimized 3D molecular structures by applying the

OPLS3e force ﬁeld. The pK_acalculation using Jaguar were based on the

protonated form of the investigated structures. The hydrogen atom

bound to the N3 atom in each molecule was designated as the basic site

for the pK_acalculations. Jaguar oﬀers a workﬂow calculation based on

DFT (density functional theory) that uses a combination of ab initio

quantum chemistry and a continuum treatment of solvation as well as

empirical corrections. Calculations were done using the B3LYP/6-

31G** level of theory. Free energy of solvation in water was calculated

through single-point calculations of the gas-phase-optimized geometry

using the Poisson−Boltzmann solvation model.

Docking Studies. For the docking studies, the co-crystal structures

with best resolution for MT1 (6ME2: R = 2.8 Å) and MT2 (6ME6: R =

2.8 Å) were employed. These structures were prepared using the

Protein Preparation Wizard⁶⁹Workﬂow provided in the MAESTRO

̈

r.2019-1 software (Schrodinger, LLC, New York, NY), using the

OPLS3e force ﬁeld. The structures were ﬁrst ﬁxed by assigning bond

orders, adding hydrogens, and zero-order bonds to metals, creating

disulﬁde bonds, ﬁlling missing side chains with Prime, deleting water

molecules >5 Å from the het group, and, with less than three H-bonds

to non-waters, generating protonation states at pH 7.0 2.0 using

Epik,⁷⁰and optimizing the H-bond network using PROPKA⁷¹with a

pH of 7.0. All the residue rotamers were retained from the PDB ﬁles,

and the histidine residues were treated in their neutral form with

protonation at N_ε. The two-dimensional structures of each ligand

(compounds 10b and 15a) were built using CANVAS (Canvas,

̈

Schrodinger, LLC, New York, NY, 2019) and exported to MAESTRO

r.2019-1. Then, LigPrep was used to generate all possible stereoisomers,

tautomers, and accurate energy-minimized 3D molecular structures

applying the OPLS3e force ﬁeld. The protonation states of the ligands

at pH 7.0 2.0 were determined using Epik. All generated states for

compounds 10b and 15a were ﬁnally docked in the corresponding

structures for MT1 (6ME2) and MT2 (6ME6) using the standard

precision mode (SP mode) in Glide,⁷²with no scaling factors applied to

the van der Waals radii and all remaining parameters in Glide were set to

default options. The receptor grid of the complex was generated with

GLIDE r.2019-4, and grid boxes were deﬁned as a 10 × 10 × 10 Å³

region centered at the co-crystal ligands. Two constraints were included

in the docking protocol: one restricting the poses to those in which the

ligands exhibited hydrogen-bond (H-bond) interactions with the

residues Gln181 and Gln194 in MT1 and MT2, respectively, and an

aromatic ring positional constraint based on the core of the co-crystal

ligands. The top-ranked pose for compounds 10b and 15a, according to

Glide, was selected for further analysis. The Glide Score uses an

empirical scoring function that approximates the ligand binding free

energy with the protein by using force ﬁeld-based components and

additional terms that account for solvation and repulsive interactions.⁷²

Absorption, Distribution, Metabolism, and Excretion

(ADME) Assays. A series of ADME assays was performed with

selected compounds to determine their aqueous solubility, cellular

permeability, hepatocyte and microsomal metabolism, cytochrome

P450 (CYP) inhibition, and plasma protein binding in diﬀerent species.

The experiments were performed at Euroﬁns Pharma Discovery (Celle-

́

Levescault, France).

1920

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

Aqueous solubility was determined in a phosphate-buﬀered saline

(PBS, pH 7.4) using the shake-ﬂask method.⁷³Compounds were

diluted in DMSO and transferred to the assay medium. Samples were

incubated at room temperature for 24 h, two independent

determinations in duplicate, and analyzed by HPLC-UV/VIS. Aqueous

solubility was determined as the ratio between the peak area from the

test sample and a calibration standard (60% methanol and 40% water).

When the solubility was greater than the maximum assay concentration,

the results were reported as 200 μM.

independent determinations. UV spectra (250−450 nm) were

recorded, and the aqueous pK_awas calculated using the Yasuda−

Shedlovsky extrapolation.⁸¹

In Vivo PK. Pharmacokinetics (PK) experiments were performed in

CD-1 mice, Wistar Han, and SD rats at 1 mg/kg i.v. and at 10 or 30 mg/

kg p.o. (SD only) for selected compounds, n = 2 per group. PK

experiments were carried out at WuXi Apptec (Shanghai, China). The

plasma concentrations of the compounds were monitored over a 24 h

period after dosing. The PK parameters were calculated using the non-

compartment model in WinNonlin v6.3 software. The data are reported

as mean SEM.

The brain concentration and B/P ratio were determined for 10b,

15a, and 19a following p.o. administration to Wistar (100 mg/kg) and

SD (10 and 30 mg/kg) rats. These data were used to assess brain

exposure and to aid in the interpretation of the FOB (Wistar rats) and

HFD (SD rats) studies. Agomelatine and ramelteon were administered

to Wistar Han at 100 mg/kg p.o. as comparators. Brain and blood

samples were collected at 0.37, 2, and 6 h (Wistar Han) or 0.5, 2, and 8 h

(SD) post dose, n = 2 (SD) or 3 (Wistar) per time point. The selection

of the time points was based on baseline oral PK data performed in each

rat strain. The B/P ratios were calculated based on the brain and plasma

concentrations. The data are reported as mean SEM at each time

point.

The concentrations of the test compounds in the plasma and brain

samples were determined using the SCIEX API 4500 HPLC-MS/MS

system, with an electrospray ionization source and triple quadrupole

analyzer. Chromatographic separation was achieved with a C18 column

(2.6 μm, 50 mm × 3.00 mm), a gradient of 98% solvent A to 100%

solvent B (solvent A: water with 0.1% formic acid; solvent B:

acetonitrile with 0.1% formic acid) at the ﬂow rate of 0.7 mL/min.

PK data for melatonin after oral (10 mg/kg) and intravenous (5 mg/

kg) administration were obtained from the literature.^43,44The results

presented as graphs in the corresponding publications were extracted

using the DigitizeIt v2.3 (Braunschweig, Germany).

HFD Rat Model. The eﬀect of repeated administration of the

melatonin receptor agonists 10b, 15a, and 19a on metabolic parameters

was assessed in SD rats maintained on HFD. Dapagliﬂozin was used as a

comparator since it reduces blood glucose levels and body weight in rat

DIO models.⁴⁷The detailed study timeline is shown in Figure S6. The

HFD and oral glucose challenge experiments were carried out at

RenaSci Ltd. (Nottingham, United Kingdom).

All animals had free access to a high-fat pelleted diet since their

arrival (D12451, 45% energy as fat, Research Diets Inc., Brunswick,

New Jersey, USA), except for 10 animals placed on standard chow

(Envigo Teklad 2018). Animals were maintained on a reversed-phase

light−dark cycle (lights oﬀ for 12 h from 11:30 to 23:30), during which

the room was illuminated by red light. Animals were kept under these

conditions during a 4 week acclimatization period (day −32 to −2). On

the fourth week of acclimatization, the rats were handled for 5 days and

then underwent a 3 day baseline run-in procedure, during which they

were weighed and dosed once a day with the vehicle. The amount of

food present in each cage was determined on the ﬁrst and last day of the

baseline period.

After the baseline period, the animals were allocated into nine

weight-matched groups (SAS/QC OPTEX), n = 10 per group,

according to the diet and test compound treatments as follows: (A)

standard chow diet and vehicle; (B) HFD and vehicle; (C) HFD and

dapagliﬂozin at 5 mg/kg; (D) HFD and 15a at 10 mg/kg; (E) HFD and

15a at 30 mg/kg; (F) HFD and 10b at 10 mg/kg; (G) HFD and 10b at

30 mg/kg; (H) HFD and 19a at 10 mg/kg; and (I) HFD and 19a at 30

mg/kg. Treatments were administered p.o. by gavage once daily for

almost 10 weeks (66 days). All dosing was timed so that the midpoint of

dosing was 1 h (i.e., 10:30 h) before the onset of the dark phase (11:30

h). This approach was taken to mimic the article by Prunet-Marcassus

et al.⁸²During the study, rats were weighed (to the nearest 0.1 g) daily

at the time of dosing.

Cellular permeability was determined using Caco-2 cells.⁷⁴The

compounds were added (10 μM) to the apical (A) or to the basolateral

(B) side, and their concentrations in the donor and in the receiver

compartments were measured following incubation for 40 min (B-A) or

60 min (A-B) at 37 °C, two independent determinations in duplicate.

Samples were analyzed by HPLC-MS/MS. The integrity of the

monolayer was veriﬁed by the Lucifer yellow rejection assay. The

apparent permeability coeﬃcient (P_appat pH 6.5/7.4) was calculated

based on the compound concentration.

Microsomal metabolic stability was determined using HLM.⁷⁵

Compounds (100 nM) were incubated with HLM (0.1 mg of protein/

mL) at 37 °C, once in duplicate, for 0, 15, 30, 45, and 60 min. Samples

were analyzed by HPLC-MS/MS. The in vitro half-life (t_1/2) and

microsomal intrinsic clearance (CL_int) of each compound were

calculated based on their disappearance rate (k, slope of the natural

logarithm of concentration versus time curve) and assuming ﬁrst-order

reaction kinetics, using the formulas t_1/2= 0.693/−k and CL_int= 0.693/

t_1/2× microsomal protein concentration.

Hepatocyte metabolic stability was determined in SD rat, CD-1

mouse, beagle dog, and human cryopreserved hepatocytes.⁷⁵

Compounds (1 μM) were incubated with the hepatocytes (0.7 × 10⁶

viable cells/mL) at 37 °C, for 0, 0.5, 1, 1.5, and 2 h. The human, mouse,

and rat hepatocyte stability experiments were conducted as two

independent determinations in duplicate. The dog hepatocyte stability

assay was performed once in triplicate. Samples were analyzed by

HPLC-MS/MS. The in vitro half-life (t_1/2) and microsomal intrinsic

clearance (CL_int) of each compound was calculated based on their

disappearance rate (k, slope of the natural logarithm of concentration

versus time curve) and assuming ﬁrst-order reaction kinetics, using the

formulas t_1/2= 0.693/−k and CL_int= 0.693/t_1/2× number of cells per

μL.

CYP inhibition experiments were performed for isoforms 1A2, 2B6,

2C8, 2C9, 2C19, 2D6, and 3A4 using recombinant human enzymes and

ﬂuorometric quantiﬁcation. Compounds (10 μM) were incubated for

20−50 min, once in duplicate, at 37 °C, with each isoform and speciﬁc

substrates.^76,77The substrates were 7-methoxy-4-triﬂuoromethylcou-

marin for 2C9 and 2D6, 7-ethoxytriﬂuoromethyl coumarin for 2B6, 3-

cyano-7-ethoxycoumarin for CYP1A2 and 2C19, dibenzylﬂuorescein

for 2C8, 7-benzyloxy-4-triﬂuoromethylcoumarin, and benzyloxyresor-

uﬁn for 3A4.

The plasma protein binding for human and SD rat was determined

using the equilibrium dialysis method.⁷⁸The compounds (10 μM)

were mixed with plasma and transferred to the donor side of the dialysis

plate. The samples were incubated for 4 h, at 37 °C, once in duplicate,

and analyzed by HPLC-MS/MS. The fraction unbound (f_u) was

calculated as the ratio of the compound concentration on the receiver

and the donor side of the membrane.

The distribution coeﬃcient (logD) was determined in n-octanol/

aqueous phosphate buﬀer (100 mM, pH 7.4) using the shake ﬂask

method.⁷⁹Compounds were incubated for 1 h, two independent

determinations in duplicate, under constant shake (880 rpm), and at

room temperature. After the incubation, samples of each layer were

analyzed by HPLC-MS/MS. LogD was determined as the log10 of the

compound amount in the organic phase over the aqueous phase.

The acid dissociation constant (pK_a) was determined using the

multi-wavelength spectrophotometric method in a Sirius T3 automatic

titrator (Sirius Analytical Instruments Ltd., UK).⁸⁰Compounds were

diluted in a solution with 20% (v/v) ionic strength adjusted water (0.15

M KCl) and 80% (v/v) methanol. The samples were pre-acidiﬁed to pH

2.0 with an HCl solution (0.5 M) and titrated to pH 12.0 with a KOH

solution (0.5 M), at 25 °C. The experiment was performed as two

FI was determined twice weekly, which includes day 1 of compound

dosing. Water intake was not monitored. During the baseline and drug

treatment periods, all animals were examined every day immediately

prior to dosing and at the end of the dosing period. Any overt

1921

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

behavioral/physiological eﬀects or other relevant observations

regarding the condition of the animals were recorded. On days 23

and 51 (weeks 4 and 8), plasma insulin and glucose were assessed after a

4 h fast. Blood sampling was organized such that it coincided when rats

were placed in a clean cage. On blood sampling days, the rats were

transferred into a clean cage (no food) to a timed schedule 2 h prior to

dosing (also to a timed schedule). Two hours after dosing (i.e., 4 h

fasted), a blood sample was taken via venipuncture. The blood samples

were spun in a cooled centrifuge and plasma (1 aliquot) stored frozen

(−80 °C) until analyzed for glucose and insulin. Blood sampling

occurred during the dark phase of the light/dark cycle under red light.

The removal of food began at approximately 07:30 h, ﬁnishing at

approximately 09:45 h. Accordingly, dosing was timed from 09:30 h to

11:45 h and blood sampling from 11:30 h to 13:45 h. The food from the

original hopper was then transferred to the new cage. On week 9 (two

cohorts, days 58 and 59), an OGTT was performed after a 16 h fast, as

per described below, and plasma was collected to quantify ALT and

AST levels.

The study was terminated on week 10 (two cohorts, days 65 and 66).

Rats were not dosed on the ﬁnal day. After the ﬁnal reading, the animals

were humanely sacriﬁced by CO₂asphyxiation with conﬁrmation of

death by cervical dislocation. Animals’ whole liver, gastrocnemius,

brown adipose tissue, and epididymal fat pads were collected, weighed,

and processed. Two sections of the liver right lobe and a portion (100−

200 mg) of the liver caudate lobe were stored at −80 °C for triglycerides

and glycogen content quantiﬁcation, respectively. The gastrocnemius

was frozen stored (−80 °C) for the determination of glycogen content.

The left liver lobe (ﬁxed in 10% formalin and taken to wax block) was

sectioned and stained with hematoxylin and eosin stain for the scoring

of liver steatosis.

The remaining carcass was weighed and used for determination of

body composition. Frozen (−20 °C) rat carcasses were individually

milled using a Retsch SM2000 Laboratory cutting mill (Cristison

Scientiﬁc Equipment Ltd., Gateshead, UK) precooled using solid

carbon dioxide. The milled carcass homogenate was allowed to reach

room temperature before carcass composition was determined using

the FoodScan near-infrared analyzer (Foss, UK).

Plasma Analysis. Plasma glucose (ThermoFisher, Inﬁnity Glucose

hexokinase liquid stable reagent, kit TR15421) and insulin (Mercodia,

Rat insulin ELISA, kit 10-1251) were quantiﬁed using commercial kits.

Plasma ALT (ALTL 04718569 190) and AST (ASTL 04657543 190)

were measured using a Cobas C111 clinical analyzer with associated

PreciControl quality control reagents Multi 1 05947626 190 and Multi

2 05947774 190. HOMA-IR was calculated as (Insulin ng/mL ×

Glucose mM)/(14.1 × 0.0455).⁸³

frozen (−80 °C) and subsequently assayed for glucose and insulin. In

the case of the B1 sample, ALT and AST were measured, and HOMA-

IR was calculated based on glucose and insulin concentrations. Area

under the curve from 0 to 60 min (AUC_0‑60min) and 0 to 120 min

(AUC_{0‑120 min}) was calculated and analyzed similarly. Total AUC was

calculated using the trapezoidal rule, depicted below. The AUCs above

baseline (AUC_B), i.e., above B2 sample values, were calculated as

AUC_120min− 2t_0hand AUC_60min− t_0h:

1

AUC_120min

=

(t_0min+ 2t_10min+ 2t_20min+ 4t_30min+ 9t_60min

12

+ 6t_120min

)

1

2

AUC_60min= AUC_120min

−

(t_60min+ t_120min)

Tissue Analyses. Animals from standard chow diet and vehicle

(Group A), HFD and vehicle (Group B), HFD and dapagliﬂozin

(Group C), and the two HFD and 10b groups (Groups F and G) were

selected for additional tissue analyses. The liver triglycerides were

quantiﬁed using a Cobas C111 clinical analyzer TRIGL 04657594 190.

Glycogen quantiﬁcation in liver and gastrocnemius samples was

performed using commercial kits (Thermo BioVision, Glycogen

Assay Kit, K646-100). The scoring of liver steatosis was conducted

by a pathologist following a method previously described.^84,85The

parameters analyzed were as follows: steatosis (0−3); lobular

inﬂammation (0−3); and hepatocellular ballooning (0−2). The total

“NAFLD activity score” (NAS) score was calculated as the sum of all

parameters (0−8).

Statistical Analysis. The inhibition constant (K_i), half-maximal

inhibitory concentration (IC₅₀) and half-maximal eﬀective concen-

tration (EC₅₀) were determined by nonlinear regression using the Hill

equation curve ﬁtting in GraphPad Prism v8.4. The PK parameters were

calculated using WinNonlin 6.3 and non-compartment model. Analysis

of the HFD study was performed using SAS v9.3 and GraphPad Prism

v8.4.

The BW of chow and HFD animals before allocation to treatment

groups was analyzed by t test. Daily BW and BW gain during the study

were analyzed by one-way ANCOVA with BW on day 1 as the covariate.

FI and average daily FI were analyzed by one-way ANCOVA with

average FI during the baseline period as the covariate. Compounds 10b,

15a, and 19a were compared with the HFD vehicle group by Williams’

test. Dapagliﬂozin was compared with the HFD vehicle group by

Student’s t test. Plasma samples from days 23 and 51 and OGTT data

(plasma glucose and insulin concentrations and AUC and AUC_Band

OGTT. The OGTT study was performed on days 58 and 59

(approximately 1 week prior to termination). Due to the size of the

study, this assay was performed as two cohorts with procedures

separated by a 24 h period. The day prior to the OGTT study, the rats

were deprived of food overnight (a total of 16 h before initial blood

sample collection in the OGTT). The food was removed and weighed,

but free access to water was continued. The OGTT was performed the

following day. The animals were moved to a separate room, which was

maintained under normal lighting and contained heated chambers that

were used to aid the blood-sampling procedure. On the morning of the

test, the animals were cannulated (lateral tail vein) beginning at

approximately 08:00 h. When all rats had been cannulated, and the

cannula checked, a blood sample was collected from each rat (Baseline

1 (B1) time point). Animals were then dosed with vehicle or drug

starting at 10:30 h to a timed schedule (60 min before the glucose

challenge). Prior to the glucose challenge, an additional blood sample

was taken (Baseline 2 (B2) time point). Therefore, baseline blood

samples were taken both before drug treatment and before the glucose

challenge at 0 min. Animals were then dosed orally with ^D-glucose (2 g/

kg) and further blood samples taken 10, 20, 30, 60, and 120 min later.

Between blood sampling, the animals were returned to the home cage

with free access to water (but not food). Blood samples were collected

into lithium heparinized tubes (Sarstedt Microvette CB300LH) at each

time point and plasma separated by centrifugation in a cooled

centrifuge to produce a single or two aliquots of plasma, which was

ALT and AST levels) were analyzed by one-way or two-way ANOVA

̌

́

followed by the Holm−Sidak’s multiple comparison test. Body

composition was analyzed by robust regression of data with treatment

as a factor and BW on day 1 as a covariate. Tissue weights were analyzed

by two-way ANCOVA with treatment and cohort as factors and BW on

day 1 as a covariate. To determine more than expected eﬀects on tissue

weights, any treatment eﬀects on BW were analyzed by two-way

ANCOVA with treatment and cohort as factors and terminal BW as a

covariate. The analysis of liver triglycerides and liver and muscle

̌

́

glycogen was done by one-way ANOVA followed by the Holm−Sidak’s

multiple comparison test. Liver histology data (total NAS, steatosis,

lobular inﬂammation, and hepatocellular ballooning scores) was

̌

́

analyzed by one-way ANOVA followed by Holm−Sidak’s post hoc

test to compare each compound to the HFD Vehicle and Chow Vehicle

groups. A p-value less than 0.05 was considered signiﬁcant.

Early Safety and Metabolic Proﬁling. The eﬀects of 10b were

investigated in the in vivo functional observational battery.⁸⁶Wistar

(Han) rats (Elevage Janvier, Le Genest-Saint-Isle, France), n = 3 per

group, were dosed with compound 10b p.o. by gavage at 25, 50, and 100

mg/kg. Animals observed in simultaneous comparison with a control

group given vehicle (non-blind conditions for the control group)

immediately before the test. Behavioral modiﬁcations, physiological

and neurotoxicity symptoms, rectal temperature, and pupil diameter

were recorded according to a standardized observation grid containing

the items: death*, convulsions*, tremor*, Straub tail*, altered activity,

1922

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

jumping*, abnormal gait* (rolling, tiptoe), motor incoordination*,

altered abdominal muscle tone, loss of grasping, akinesia, catalepsy, loss

of traction, loss of balance*, fore-paw treading*, writhing*,

piloerection*, stereotypies* (sniﬃng, chewing, head movements),

head-twitches*, scratching*, altered respiration*, aggression*, altered

fear/startle, altered reactivity to touch, ptosis, exophthalmia, loss of

righting reﬂex, loss of corneal reﬂex, analgesia, defecation/diarrhea,

salivation, lacrimation, rectal temperature (hypothermia/hyperther-

mia), and pupil diameter (myosis/mydriasis). Observations were

performed 15, 30, 60, 120, and 180 min after administration of the test

substance and 24 h later. The symptoms marked (*) were observed

continuously from 0 to 15 min after administration. The in vivo

functional observation study was conducted at Porsolt (Le Genest-

Saint-Isle, France).

The bacterial reverse mutation (Ames) test was performed based on

the original method.⁸⁷The compound (10, 25, 50, and 100 μM) was

incubated with Salmonella typhimurium strains TA98, TA100, TA1535,

and TA1537, for 96 h, at 37 °C, in the presence and absence of rat liver

S9, n = 48 wells per concentration. Prior to the Ames test, the absence of

bacterial cytotoxicity (considered relevant if less than 60% of control

growth was observed in the presence of the compound) was shown at

concentrations ranging from 600 M to 100 μM. The number of wells in

which bacterial growth was observed due to the reversion of the

histidine mutation (as judged by the ratio of OD₄₃₀/OD₅₇₀being

greater than 1.0) was determined. The signiﬁcance of the positive

counts was assessed using the one-tailed Fisher’s exact test to compare

each treatment concentration to control. Diﬀerences were signiﬁcant if

p < 0.05. The Ames study was run at Euroﬁns Pharma Discovery (Celle-

samples were solubilized in CDCl₃or DMSO-d₆, and analysis time

varied accordingly for each sample.

General Procedure for the Synthesis of 2-Alkoxybenzimida-

zole Derivatives: 8a−i and 14a−b. Preparation of Compound 8a.

As previously described by Depreux et al.,³⁸to a mixture of 2-ﬂuoro-4-

methoxy-1-nitro-benzene 1a (1.46 mol, 1.0 eq) and ethane-1,2-diamine

(7.30 mol, 489 mL, 5.0 eq) was added CuBr₂(511 mmol, 0.35 eq) in an

ice-water bath, and the resulting reaction was stirred at 60 °C for 2 h.

The reaction mixture was diluted with H₂O (1500 mL), and the

precipitate was collected by ﬁltration to give pure product 2a N-(5-

methoxy-2-nitrophenyl)ethane-1,2-diamine (91% yield) as a yellow

solid. The product was used into next step without further puriﬁcation.

In a reactor equipped with a reﬂux condenser, magnetic stirring, and

heating, a solution of the previously obtained product 2a (1.33 mol, 1.0

eq) in EtOH (1300 mL) was added dropwise to Ac₂O (1.33 mol, 124

mL, 1.0 eq) at 20 °C. The reaction was heated to a temperature of 80 °C

and kept under reﬂux for 1 h. Volatile solvent was removed and diluted

with MTBE (2000 mL). The resultant mixture was stirred at 20 °C for

20 min and ﬁltered, and the ﬁlter cake was dried under reduced pressure

to give the intermediate N-[2-(5-methoxy-2-nitrophenylamino)ethyl]-

acetamide 3a (320 g, 95.3% yield) as a yellow solid, which was used into

1

next step directly without further puriﬁcation. H NMR (400 MHz,

CDCl₃) δ 1.80 (s, 3H), 3.25 (q, J = 6.6 Hz, 2H), 3.40 (q, J = 6.6 Hz,

2H), 3.87 (s, 3H), 6.27 (d, J = 6.0 Hz, 1H), 6.53 (s, 1H), 8.00 (d, J = 6.0

Hz, 1H), 8.14 (br s, 1H), 8.47 (br s, 1H).

To a suspension of the obtained compound 3a (592 mmol, 1.0 eq) in

MeOH (1500 mL) was added Pd/C (15.0 g, 10% purity) under a N₂

atmosphere. The suspension was degassed and purged with H₂three

times. The mixture was stirred under H₂(50 psi) at 30 °C for 4 h. The

suspension was ﬁltered through a pad of celite, and the ﬁlter cake was

washed with MeOH (500 mL x 3). The intermediate N-[2-(2-amino-5-

methoxyphenylamino)ethyl]acetamide 5a (91% yield) was obtained as

a brown solid, which was used to next step directly without further

puriﬁcation. ¹H NMR 400 MHz DMSO δ 1.80 (s, 3H). 3.25 (q, J = 6.6

Hz, 2H), 3.40 (q, J = 6.6 Hz, 2H), 3.87 (s, 3H), 6.27 (d, J = 6.0 Hz, 1H),

6.53 (s, 1H), 8.00 (d, J = 6.0 Hz, 1H), 8.14 (br s, 1H), 8.47 (br s, 1H).

In a 500 mL reactor, a mixture of 5a (537 mmol, 1.0 eq),

tetraethoxymethane (1.34 mol, 2.5 eq), and AcOH (0.1 eq) was heated

to 80 °C and kept at this temperature for 30 min. Then, the reaction

mixture was cooled to room temperature and concentrated under

reduced pressure. The crude product was puriﬁed by column

chromatography to give an oﬀ-white solid, which was washed with

MTBE (400 mL) for further puriﬁcation. Compound N-[2-(2-ethoxy-

6-methoxybenzoimidazol-1-yl)ethyl]acetamide (8a) was obtained as a

white solid (62% yield). ¹H NMR (300 MHz, CDCl₃) δ ppm 1.47 (t, J =

7.1 Hz, 3 H); 1.93 (s, 3 H); 3.57−3.61 (m, 2 H); 3.84 (s, 3 H); 4.11−

4.14 (m, 2 H); 4.57 (q, J = 7.1 Hz, 2 H); 5.78 (br s, 1 H); 6.73−6.80 (m,

2 H); 7.41 (d, J = 8.6 Hz, 1 H). ¹³C NMR (75 MHz, CDCl₃) δ ppm

14.73; 23.14; 38.94; 41.27; 55.99; 66.32; 93.46; 109.34; 117.93;

133.59; 134.24; 155.58; 156.62; 170.63. HRMS (ESI) m/z calculated

C₁₄H₁₉N₃O₃: 277.1426, found [M + H]⁺278.1516.

́

Levescault, France).

The metabolic proﬁling of 10b was investigated using human

cryopreserved hepatocytes. The compound was incubated at 10 μM,

1.0 × 10⁶viable cells/ mL, at 37 °C, for 120 min, n = 1. Samples were

analyzed by HPLC-MS/MS. MetaboLynx software was employed for

data processing, and the structures of the metabolites were elucidated

based on the characteristics of their ion masses (MS spectrum) and

their secondary fragmentation products (MS²). The metabolic proﬁling

experiment was conducted at WuXi Apptec (Shanghai, China).

Chemical Synthesis. All starting materials were obtained from

commercial suppliers and used without further puriﬁcation. Solvents for

chemical transformations, work-up, and chromatography were

purchased from Sigma-Aldrich at HPLC grade. The progress of all

reactions was monitored by analytical thin-layer chromatography,

which was performed on silica-gel Silicagel 60 F₂₅₄plates (Merck).

Visualization was accomplished with UV light. Chromatographic

puriﬁcations were carried out through a silica gel column using a

Biotage Isolera Dalton 2000 automated system in gradients of

dichloromethane and methanol.

All the test compounds exhibited >95% purity as determined by

HPLC. HPLC analysis was performed using a 1260 Inﬁnity II LC

Agilent system with fraction collector equipped with a C18 column (5

μm, 4.6 mm × 250 mm) with a gradient of 95% solvent A to 75% solvent

B (solvent A: 10% aqueous methanol solution; solvent B: acetonitrile +

0.01% formic acid), ﬂow rate = 1.0 mL/min, and UV detection at 254

nm. Alternatively, LC-MS analysis was performed using Agilent 1200 &

6110 instrument equipped with a Halo-C18 (3.0 × 30 mm, 2.7 μm)

using a gradient of 5−95% B in 3 min (solvent A: 0.037% triﬂuoroacetic

acid in water; solvent B: 0.018% triﬂuoroacetic acid in acetonitrile 10%

aqueous methanol solution), ﬂow rate = 0.8 mL/min, and detection

methods were diode array (DAD) and evaporative light scattering

(ELSD) detection as well as positive electrospray ionization. High-

resolution mass spectrometry (HRMS) data (m/z) were obtained using

a UHPLC Agilent 1290 Inﬁnity with MS Agilent Q-TOF 6540 UHD

Accurate-Mass system with an electrospray ionization source (Agilent

Jet Stream, AJS) and time of ﬂight (TOF) analyzer.

Preparation of Intermediates 3b−i. Compounds 3b−i were

prepared according to the same route as described for the compound

3a using R³−COCl instead of Ac₂O.

Preparation of Compound 8b. N-[2-(2-Ethoxy-6-methoxybenzoi-

midazol-1-yl)ethyl]propionamide (8b): Compound 8b was prepared

(white solid, 57% yield) from 3b according to the same route as

described for the compound 8a. ¹H NMR (500 MHz, CDCl₃) δ ppm

1.09−1.12 (m, 3H); 1.47 (t, J = 7.1 Hz, 3 H); 2.13 (q, J = 7.5 Hz, 2 H);

3.59 (q, J = 5.9 Hz, 2 H); 3.83 (s, 3 H); 4.12 (t, J = 5.8 Hz, 2 H); 4.55 (q,

J = 7.1 Hz, 2 H); 5.64 (br s, 1 H); 6.72 (d, J = 2.4 Hz, 1 H); 6.79 (d, J =

8.6, 2.4 Hz, 1 H); 7.41 (d, J = 8.6 Hz, 1H). HRMS (ESI) m/z calculated

C₁₅H₂₁N₃O₃: 291.1583; found [M + H]⁺292.1659.

Preparation of Compound 8c. N-[2-(2-Ethoxy-6-methoxybenzoi-

midazol-1-yl)ethyl]butyramide (8c): Compound 8c was prepared

(white solid, 53% yield) from 3c according to the same route as

described for the compound 8a. ¹H NMR (500 MHz, CDCl₃) δ ppm

0.91 (t, J = 7.4 Hz, 3 H); 1.46 (t, J = 7.1 Hz, 3 H); 1.57−1.67 (m, 2 H);

2.06−2.09 (m, 2 H); 3.60 (q, J = 6.0 Hz, 2 H); 3.83 (s, 3 H); 4.08−4.14

(m, 2 H); 4.54 (q, J = 7.1 Hz, 2 H); 5.64 (br s, 1 H); 6.72 (d, J = 2.7 Hz,

All ¹H and ¹³C NMR spectra were recorded using a Bruker

spectrometer (model AvanceIII 400 or 500 MHz) at 25 °C. Chemical

shifts (δ) are given relative to SiMe₄and referenced to the residual

solvent signal. Chemical shifts are reported in ppm relative, and

coupling constants (J) are reported in Hz to the nearest 0.5. All the

1923

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

1 H); 6.76−6.79 (m, 1 H); 7.40 (d, J = 8.6 Hz, 1 H). HRMS (ESI) m/z

calculated C₁₆H₂₃N₃O₃: 305.1739; found [M + H]⁺306.1836.

Preparation of Compound 8d. N-[2-(2,6-Dimethoxybenzoimida-

zol-1-yl)ethyl]acetamide (8d): Compound 8d was prepared (white

solid, 64% yield) from 3a according to the same route as described for

the compound 8a, using tetramethoxymethane instead of tetraethoxy-

ethane. ¹H NMR (300 MHz, CDCl₃) δ ppm 1.92 (s, 3 H); 3.57 (q, J =

5.9 Hz, 2 H); 3.83 (s, 3 H); 4.06−4.12 (m, 2 H); 4.13 (s, 3H); 5.82 (br

s, 1 H); 6.72 (d, J = 2.7 Hz, 1 H); 6.75−6.81 (m, 1 H); 7.40 (d, J = 8.6

Hz, 1 H). ¹³C NMR (75 MHz, CDCl₃) δ ppm 23.13; 38.96; 41.24;

55.99; 57.02; 93.45; 109.32; 118.16; 133.82; 134.56; 155.61; 157.34;

170.68. HRMS (ESI) m/z calculated C₁₃H₁₇N₃O₃: 263.1270; found

[M + H]⁺264.1374.

Preparation of Compound 8e. Cyclopropanecarboxylic acid [2-(2-

ethoxy-6-methoxybenzoimidazol-1-yl)ethyl]amide (8e): Compound

8e was prepared (white solid, 58% yield) from 3d according to the same

route as described for the compound 8a. ¹H NMR (500 MHz, CDCl₃)

δ ppm 0.71−0.74 (m, 2 H); 0.97−0.99 (m, 2 H); 1.22−1.25 (m, 1 H);

1.45 (t, J = 7.1 Hz, 3 H); 3.61 (q, J = 5.9 Hz, 2 H); 3.83 (s, 3 H); 4.10 (t,

J = 5.6 Hz, 2 H); 4.51 (q, J = 7.0 Hz, 2 H); 5.96 (br s, 1 H); 6.71 (d, J =

2.3 Hz, 1 H); 6.78 (dd, J = 8.7, 2.4 Hz, 1 H); 7.40 (d, J = 8.5 Hz, 1 H).

HRMS (ESI) m/z calculated C₁₆H₂₁N₃O₃: 303.1583; found [M + H]⁺

304.1678.

Preparation of Compound 8f. Cyclobutanecarboxylic acid [2-(2-

ethoxy-6-methoxybenzoimidazol-1-yl)ethyl]amide (8f): Compound 8f

was prepared (white solid, 50% yield) from 3e according to the same

route as described for the compound 8a. ¹H NMR (500 MHz, CDCl₃)

δ ppm 1.47 (t, J = 7.1 Hz, 3 H); 1.83−1.97 (m, 2 H); 2.07−2.11 (m,

2H); 2.19−2.25 (m, 2H); 2.85−2.92 (m, 1H); 3.60 (q, J = 5.9 Hz, 2 H);

3.83 (s, 3 H); 4.11 (t, J = 5.8 Hz, 2 H); 4.55 (q, J = 7.2 Hz, 2 H); 5.54

(br, 1 H); 6.71 (d, J = 2.3 Hz, 1 H); 6.78 (dd, J = 8.7, 2.4 Hz, 1 H); 7.41

(d, J = 8.70 Hz, 1 H). HRMS (ESI) m/z calculated C₁₇H₂₃N₃O₃:

317.1739; found [M + H]⁺= 318.1855.

Preparation of Compound 8g. Cyclopentanecarboxylic acid [2-(2-

ethoxy-6-methoxybenzoimidazol-1-yl)ethyl]amide (8g): Compound

8g was prepared (white solid, 51% yield) from 3f according to the same

route as described for the compound 8a. ¹H NMR (500 MHz, CDCl₃)

δ ppm 1.46 (t, J = 7.1 Hz, 3H); 1.50−1.53 (m, 2 H); 1.67−1.76 (m, 6

H); 2.35−2.43 (m, 1 H); 3.60 (q, J = 5.8 Hz, 2 H); 3.82 (s, 3 H); 4.10 (t,

J = 5.8 Hz, 2 H); 4.54 (q, J = 7.2 Hz, 2 H); 5.62 (br s, 1 H); 6.71 (d, J =

2.3 Hz, 1 H); 6.77 (dd, J = 8.6, 2.4 Hz, 1 H); 7.40 (d, J = 8.7 Hz, 1 H).

HRMS (ESI) m/z calculated C₁₈H₂₅N₃O₃: 331.1896; found [M + H]⁺

= 332.1987.

(br s, 1 H); 6.68 (d, J = 2.3 Hz, 1 H); 6.78 (dd, J = 8.6, 2.4 Hz, 1 H); 7.42

(d, J = 8.6 Hz, 1 H). ¹³C NMR (75 MHz, CDCl₃) δ ppm 14.78; 23.28;

28.56; 36.86; 39.50; 56.05; 66.21; 93.90; 108.71; 118.11; 133.86;

134.22; 155.37; 156.74; 170.15. HRMS (ESI) m/z calculated

C₁₅H₂₁N₃O₃: 291.1583; found [M + H]⁺292.1699.

Preparation of Compound 14b. N-[3-(2,6-Dimethoxybenzoimida-

zol-1-yl)propyl]acetamide (14b): Compound 14b was prepared (white

solid, 59% yield) from intermediate 13 according to the same route as

described for the compound 8a, using propane-1,3-diamine instead of

ethane-1,2-diamine. ¹H NMR (300 MHz, CDCl₃) δ ppm 1.94 (s, 3H);

1.98−2.03 (m, 2 H); 3.26 (q, J = 6.72 Hz, 2 H); 3.85 (s, 3 H); 3.96−

4.00 (m, 2 H); 4.17 (s, 3H); 5.57 (br s, 1 H); 6.68 (d, J = 2.3 Hz, 1 H);

6.78 (dd, J = 8.6, 2.4 Hz, 1 H); 7.43 (d, J = 8.6 Hz, 1 H). ¹³C NMR (75

MHz, CDCl₃) δ ppm 23.26; 28.67; 37.00; 39.68; 56.03; 57.17; 93.91;

108.75; 118.19; 134.06; 155.42; 157.32; 170.21. HRMS (ESI) m/z

calculated C₁₄H₁₉N₃O₃: 277.1426; found [M + H]⁺278.1513.

Preparation of Compound 8j. N-[2-(2-Ethoxy-5-methoxybenzoi-

midazol-1-yl)ethyl]acetamide (8j): Compound 8j was prepared (white

solid, 52% yield) from 3a and 1b as starting materials according to the

1

same route as described for the compound 8a. H NMR (500 MHz,

CDCl₃) δ ppm 1.46 (t, J = 7.1 Hz, 3H), 1.90 (s, 3H), 3.55−3.59 (m,

2H), 3.83 (s, 3 H), 4.11 (t, J = 6.2 Hz, 2H), 4.56 (q, J = 7.1 Hz, 2H),

6.76 (dd, J = 2.4 Hz, J = 8.6 Hz, 1H), 7.04 (d, J = 8.6 Hz, 1 H), 7.10 (d, J

= 2.4 Hz, 1H). HRMS (ESI) m/z calculated C₁₄H₁₉N₃O₃: 277.1426;

found [M + H]⁺= 278.1531.

Preparation of Compound 4. N-[2-(6-methoxybenzoimida-

zol-1-yl)ethyl]acetamide. A mixture of the N-[2-(5-methoxy-2-

nitrophenylamino)ethyl]acetamide 3a (2.2 mmol) and formic acid

(114 mmol) in THF (12 mL) was hydrogenated under pressure (35

bar) over 10% palladium on carbon (0.05 g) at 60 °C for 8 h. After

ﬁltration and evaporation, the residue was puriﬁed by column

chromatography on silica gel to give the desired compound N-[2-(6-

methoxybenzoimidazol-1-yl)ethyl]acetamide (4) as an oﬀ-white solid

(46% yield). ¹H NMR (500 MHz, CDCl₃) δ ppm 1.96 (s, 3H), 3.63 (q,

J = 6.0 Hz, 2H), 3.87 (s, 3 H), 4.31 (t, J = 6.0 Hz, 2H), 6.86−6.91 (m,

2H), 7.61 (d, J = 8.8 Hz, 1 H), 7.67 (s, 1H). HRMS (ESI) m/z

calculated C₁₂H₁₅N₃O₂: 233.1164; found [M + H]⁺234.1259. The

chemical physical data are in agreement to those previously reported.³⁸

Preparation of Compound 6. To a solution of N-[2-(2-amino-5-

methoxyphenylamino)ethyl]acetamide 5a (0.05 mmol) in anhydrous

THF (4 mL) was added Et₃N (0.5 mL). Then, a solution of BTC (0.01

mmol) in anhydrous THF (2 mL) was added dropwise to the mixture at

0 °C and the reaction mixture was stirred at 15 °C for 12 h. The reaction

mixture was extracted with H₂O (5 mL) and EtOAc (20 mL x 3). The

organic layers were washed with brine (20 mL), dried over anhydrous

Na₂SO₄, ﬁltered, and concentrated. The residue was puriﬁed by

chromatography on silica gel (CH₂CI₂) to give compound N-[2-(6-

methoxy-2-oxo-2,3-dihydrobenzoimidazol-1-yl)ethyl]acetamide (6) as

an white solid (54% yield). ¹H NMR (500 MHz, CDCl₃) δ ppm 1.96 (s,

3H), 3.57−3.61 (m, 2H), 3.82 (s, 3 H), 4.02 (t, J = 6.0 Hz, 2H), 6.64

(dd, J = 2.4 Hz, J = 8.4 Hz, 1H), 6.66 (d, J = 2.4 Hz, 1H), 6.94 (d, J = 8.6

Hz, 1 H). HRMS (ESI) m/z calculated C₁₂H₁₅N₃O₃: 249.1113; found

[M + H]⁺250.1188.

Preparation of Compound 8h. Cyclohexanecarboxylic acid [2-(2-

ethoxy-6-methoxybenzoimidazol-1-yl)ethyl]amide (8h): Compound

8h was prepared (white solid, 55% yield) from 3g according to the same

route as described for the compound 8a. ¹H NMR (500 MHz, CDCl₃)

δ ppm 1.17−1.20 (m, 2 H); 1.28−1.36 (m, 2H); 1.46 (t, J = 7.1 Hz,

3H); 1.57−1.77 (m, 6 H); 1.94−2.00 (m, 1 H); 3.60 (q, J = 5.9 Hz, 2

H); 3.83 (s, 3 H); 4.10 (t, J = 5.7 Hz, 2 H); 4.55 (q, J = 7.2 Hz, 2 H);

5.63 (br s, 1H); 6.71 (d, J = 2.3 Hz, 1 H); 6.77 (dd, J = 8.7, 2.4 Hz, 1 H);

7.41 (d, J = 8.7 Hz, 1 H). HRMS (ESI) m/z calculated C₁₉H₂₇N₃O₃:

345.2052; found [M + H]⁺346.2131.

Preparation of Compound 7. N-[2-(2-Amino-5-

methoxyphenylamino)ethyl]acetamide 5a (6 mmol) and thiourea (6

mmol) were placed into a 10 mL sealed tube. The solid mixture was

initially heated to 120 °C for 10 min with intense vapor release and then

heated to 160 °C for 5 min with a second vapor release. Later, the

temperature was reduced to 80 °C and 15 mL of EtOH was added to

the mixture. The resultant mixture was cooled to −10 °C, and the

product was ﬁltrated and washed, sequentially, three times with 10 mL

of cold ethanol yielding 1.20 g of the product, which was used into next

step directly without further puriﬁcation. To a solution of the obtained

compound (0.18 mmol) in anhydrous acetone (2 mL) was added 0.094

mmol of K₂CO₃followed by the dropwise addition of 0.094 mmol of

methyl iodide at 0 °C. The resultant mixture was stirred at room

temperature for 1 h. Then, a second portion of K₂CO₃(0.094 mmol)

and methyl iodide (0.094 mmol) was added to the mixture and the

reaction was kept under stirring, at room temperature, overnight. Later,

the solution was concentrated under reduced pressure. The crude

Preparation of Compound 8i. 4-oxo-4H-pyran-2-carboxylic acid

[2-(2-ethoxy-6-methoxybenzoimidazol-1-yl)ethyl]amide (8i): Com-

pound 8i was prepared (white solid, 61% yield) from 3h according to

the same route as described for the compound 8a. ¹H NMR (500 MHz,

CDCl₃) δ ppm 1.42 (t, J = 7.1 Hz, 3H), 3.76−3.80 (m, 2H), 3.78 (s, 3

H), 4.20−4.23 (m, 2H), 4.54 (q, J = 7.1 Hz, 2H), 6.39 (dd, J = 2.6 Hz, J

= 5.8 Hz, 1H), 6.68 (d, J = 2.6 Hz, 1H), 6.78 (dd, J = 2.6 Hz, J = 8.8,

1H), 7.14 (d, J = 2.6 Hz, 1 H), 7.42 (d, J = 8.8 Hz, 1H), 7.59 (d, J = 5.8,

1H). HRMS (ESI) m/z calculated C₁₈H₁₉N₃O₅: 357.1325; found [M +

H]⁺358.1408.

Preparation of Compound 14a. N-[3-(2-Ethoxy-6-methoxyben-

zoimidazol-1-yl)propyl]acetamide (14a): Compound 14a was pre-

pared (white solid, 58% yield) from intermediate 13 according to the

same route as described for the compound 8a, using propane-1,3-

diamine instead of ethane-1,2-diamine. ¹H NMR (300 MHz, CDCl₃) δ

ppm 1.49 (t, J = 7.1 Hz, 3 H); 1.93−2.09 (m, 5 H); 3.25 (q, J = 6.7 Hz, 2

H); 3.85 (s, 3H); 3.97−4.02 (m, 2 H); 4.59 (q, J = 7.1 Hz, 2 H); 5.59

1924

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

product was diluted in EtOAc (10 mL) and extracted with H₂O (2 x 5

mL). The organic layers were washed with brine (10 mL), dried over

anhydrous Na₂SO₄, ﬁltered, and concentrated. The residue was puriﬁed

by column chromatography on silica gel (CH₂CI₂/MeOH = 10:1) to

aﬀord compound N-[2-(6-methoxy-2-methylsulfanylbenzoimidazol-1-

yl)ethyl]acetamide (7) (white solid, 80% yield). ¹H NMR (500 MHz,

CDCl₃) δ ppm 1.93 (s, 3 H); 2.76 (s, 3 H); 3.59−3.63 (m, 2 H); 3.85 (s,

3 H); 4.24 (t, J = 5.8 Hz, 2 H); 5.67 (br s, 1H); 6.78−6.79 (m, 1 H);

6.85 (d, J = 8.7 Hz, 1 H); 7.54 (d, J = 8.7 Hz, 1 H). HRMS (ESI) m/z

calculated C₁₃H₁₇N₃O₂S: 279.1041; found [M + H]⁺280.1148.

Preparation of Compound 9. To a solution of compound 8a

(0.46 mmol) in MeOH (5 mL) was added NaOH (0.46 mmol) and

heated at 80 °C for 8 h. After cooling to room temperature, a solution of

HCl/EtOH was added to the mixture until pH = 2−3 and it was

concentrated. The residue was puriﬁed by silica gel ﬂash column

chromatography (CH₂Cl₂/MeOH = 10:1) to aﬀord compound 2-(2-

ethoxy-6-methoxybenzoimidazol-1-yl)ethylamine (9) (white solid,

HRMS (ESI) m/z calculated C₁₄H₁₈ClN₃O₃: 311.1037; found [M +

H]⁺312.1165.

General Procedure for the Synthesis of Dihydrobenzofuran

Derivatives: 19a and 19b. Preparation of Compound 19a. The

synthetic route of the intermediate 16, tert-butyl {2-[(5-amino-2,3-

dihydro-1-benzofuran-4-yl)amino]ethyl}carbamate, was previously

described by Koike et al.³³In a 1000 mL reactor, a mixture of 16

(307 mmol, 1.0 eq), tetramethoxymethane (767 mmol, 2.5 eq), and

AcOH (30.7 mmol, 1.75 mL, 0.1 eq) was heated to 80 °C for 30 min.

The mixture was cooled to room temperature and concentrated under

reduced pressure to give a residue, which was triturated with MTBE

(500 mL) at 15 °C for 30 min. The product 17 (240 mmol, 78.2%

yield) was obtained as a white solid. ¹H NMR (400 MHz, DMSO) δ

1.29 (s, 9H), 3.18 (q, J = 6.2 Hz, 2H), 3.41 (t, J = 8.0 Hz, 2H), 3.95 (t, J

= 6.2 Hz, 2H), 4.02 (s, 3H), 4.53 (t, J = 8.6 Hz, 2H), 6.51 (d, J = 8.4 Hz,

1H), 7.01−7.08 (m, 2H).

To a solution of the intermediate 17 (210 mmol, 1.0 eq) in DCM

(350 mL) was added TFA (1.47 mol, 109 mL, 7.0 eq) at 20 °C. Then,

the mixture was stirred at 20 °C for 6 h. The resulting solution was

diluted with dichloromethane (200 mL), and the pH value was adjusted

with a 15% aqueous sodium carbonate solution to around 12. The

organic phase was separated, dried with magnesium sulfate, and

evaporated. The product 18 (150 mmol, 71.5% yield) was obtained as a

white solid and used directly in the next step without further

puriﬁcation. ¹H NMR (400 MHz, CDCl₃) δ 3.03 (t, J = 6.4 Hz, 2H),

3.44 (t, J = 4.4 Hz, 2H), 4.01 (t, J = 6.2 Hz, 2H), 4.15 (s, 3H), 4.63 (t, J =

8.6 Hz, 2H), 7.27 (d, J = 8.4 Hz, 1H), 7.67 (d, J = 8.6 Hz, 1H).

In a reactor equipped with reﬂux condenser, magnetic stirring, and

heating, acetic anhydride (165 mmol, 15.5 mL, 1.1 eq) was added

dropwise to a solution of the intermediate 18 (150 mmol, 1.0 eq) in

EtOH (200 mL) at 20 °C. The reaction was heated to 80 °C and kept

under reﬂux for 1 h. The volatile solvent was removed, diluted with

MTBE (150 mL), and stirred at 20 °C for 20 min. The resulting

suspension was ﬁltered and dried under reduced pressure to give a white

solid. The crude product was triturated with EtOH (90 mL) at 20 °C

for 30 min, and the solid was collected by ﬁltration, which was dried

under evaporator to give compound N-[2-(2-methoxy-7,8-dihydro-6-

oxa-1,3-diaza-as-indacen-1-yl)ethyl]acetamide (19a) as a yellow solid

(31.0 g, 113 mmol, 75% yield). ¹H NMR (400 MHz, CDCl₃) δ 1.95 (s,

3H), 3.43 (t, J = 8.6 Hz, 2H), 3.56 (t, J = 6.2 Hz, 2H), 4.08−4.12 (m,

5H), 4.63 (t, J = 8.6 Hz, 2H), 5.91 (br s, 1H), 6.67 (d, J = 8.4 Hz, 1H),

7.25 (d, J = 8.4 Hz, 1H). ¹³C NMR (75 MHz, CDCl₃) δ ppm 23.09;

28.04; 39.64; 42.07; 56.89; 71.45; 103.82; 106.32; 116.57; 130.78;

134.26; 156.74; 157.09; 170.65. HRMS (ESI) m/z calculated

C₁₄H₁₇N₃O₃: 275.1270; found [M + H]⁺= 276.1358.

1

56% yield). H NMR (500 MHz, CDCl₃) δ ppm 1.46 (t, J = 7.1 Hz,

3 H), 1.77−1.80 (m, 2H), 3.07 (t, J = 6.2 Hz, 2 H), 3.84 (s, 3 H), 3.99 (t,

J = 6.2 Hz, 2 H) 4.58 (q, J = 7.1 Hz, 2H), 6.74 (d, J = 2.4 Hz, 1H), 6.78

(dd, J = 2.4 Hz, J = 8.6 Hz, 1 H) 7.42 (d, J = 8.6 Hz, 1 H). HRMS (ESI)

m/z calculated C12H17N3O2: 235.1321; found [M + H]⁺236.1421.

General Procedure for the Synthesis of 5-Chlorination of 6-

Methoxybenzimidazole Derivatives: 10a−b, 15a−b. Prepara-

tion of Compound 10a. The compound N-[2-(2-ethoxy-6-methox-

ybenzoimidazol-1-yl)ethyl]acetamide 8a was synthesized as described

previously.³⁸In a 1000 mL reactor, NCS (292 mmol, 1.1 eq) was added

to a solution of the obtained compound 8a (266 mmol, 1.0 eq) in i-

PrOH (560 mL) at 20 °C in one portion. The reaction was heated to 90

°C and kept under reﬂux for 16 h. The mixture was cooled to room

temperature and concentrated under reduced pressure. The obtained

oil was dissolved in DCM (1000 mL) and washed with 10% aqueous

sodium carbonate solution (300 mL). The organic portion was dried

over magnesium sulfate and concentrated under reduced pressure to

give an oﬀ-white solid. The oﬀ-white solid was puriﬁed by column

chromatography followed by trituration with EtOH (70 mL) at 15 °C

for 20 min for three time to give target compound N-[2-(5-chloro-2-

ethoxy-6-methoxybenzoimidazol-1-yl)ethyl]acetamide (10a) as a

white solid (61% yield). ¹H NMR (500 MHz, CDCl₃) δ ppm 1.46 (t,

3 H); 1.92 (s, 3 H); 3.54−3.58 (m, 2 H); 3.92 (s, 3 H); 4.12 (t, J = 5.87

Hz, 2 H); 4.50−4.54 (m, 2 H); 5.75 (br, 1H); 6.79 (s, 1H); 7.50 (s, 1

H). ¹³C NMR (75 MHz, CDCl₃) δ ppm 14.70; 23.17; 39.02; 41.13;

56.95; 66.41; 93.00; 116.67; 118.99; 133.00; 133.82; 150.76; 157.05;

170.76. HRMS (ESI) m/z calculated C₁₄H₁₈ClN₃O₃: 311.1037; found

[M + H]⁺312.1134.

Preparation of Compound 19b. N-[2-(2-Ethoxy-7,8-dihydro-6-

oxa-1,3-diaza-as-indacen-1-yl)ethyl]acetamide (19b): Compound 19b

was prepared (white solid, 72% yield) according to the same route as

described for the compound 19a, using tetraethoxymethane instead of

tetramethoxymethane. ¹H NMR (300 MHz, DMSO-d₆) δ ppm 1.38 (t,

J = 7.1 Hz, 3 H); 1.75 (s, 3 H); 3.29−3.44 (m, 4 H); 4.00 (t, J = 6.1 Hz, 2

H); 4.38−4.57 (m, 4 H); 6.53 (d, J = 8.4 Hz, 1 H); 7.08 (d, J = 8.4 Hz,

1H); 8.09 (br s, 1H). ¹³C NMR (75 MHz, DMSO-d₆) δ ppm 14.98;

22.89; 27.77; 39.25; 42.19; 66.04; 71.33; 103.13; 107.21; 116.05;

131.01; 134.81; 156.13; 156.69; 170.03. HRMS (ESI) m/z calculated

C₁₅H₁₉N₃O₃: 289.1426; found [M + H] + 290.1551.

Preparation of Compound 10b. N-[2-(5-Chloro-2,6-dimethox-

ybenzoimidazol-1-yl)ethyl]acetamide (10b): Compound 10b was

prepared (white solid, 39% yield) from 8d according to the same

route as described for the compound 10a. ¹H NMR (400 MHz, CDCl₃)

δ 1.91 (s, 3H), 3.52−3.57 (m, 2H), 3.91 (s, 3H), 4.10−4.14 (m, 5H),

5.62 (br s, 1H), 6.80 (s, 1H), 7.52 (s, 1H). ¹³C NMR (125 MHz,

CDCl₃) δ ppm 23.16; 39.00; 41.16; 56.94; 57.21; 93.02; 116.73;

119.08; 133.20; 133.68; 150.85; 157.61; 170.81. HRMS (ESI) m/z

calculated C₁₃H₁₆ClN₃O₃: 297.0880; found [M + H]⁺298.0977.

Preparation of Compound 15a. N-[3-(5-Chloro-2-ethoxy-6-

methoxybenzoimidazol-1-yl)propyl]acetamide (15a): Compound

15a was prepared (white solid, 32% yield) from 14a according to the

same route as described for the compound 10a. ¹H NMR (400 MHz,

CDCl₃) δ 1.48 (t, J = 7.0 Hz, 3H), 1.94−2.01 (m, 5H), 3.24 (q, J = 4.8

Hz, 2H), 3.92 (s, 3H), 3.99 (t, J = 6.8 Hz, 2H), 4.57 (q, J = 6.8 Hz, 2H),

5.58 (br s, 1H), 6.71 (s, 1H), 7.52 (s, 1H). ¹³C NMR (125 MHz,

CDCl₃) δ ppm 14.74; 23.28; 28.73; 36.91; 39.70; 57.11; 66.48; 93.25;

116.79; 119.10; 132.35; 134.10; 150.67; 157.06; 170.22. HRMS (ESI)

m/z calculated C₁₅H₂₀ClN₃O₃: 325.1193; found [M + H]⁺326.1292.

Preparation of Compound 15b. N-[3-(5-Chloro-2,6-dimethox-

ybenzoimidazol-1-yl)propyl]acetamide (15b): Compound 15b was

prepared (white solid, 68% yield) from 14b according to the same route

described for 10a. ¹H NMR (500 MHz, CDCl₃) δ ppm 1.95 (s, 3H);

1.99−2.01 (m, 2 H); 3.27 (q, J = 6.8 Hz, 2 H); 3.93 (s, 3 H); 3.99 (t, J =

6.9 Hz, 2 H); 4.18 (s, 3 H); 5.54 (br s, 1 H); 6.72 (s, 1 H); 7.54 (s, 1 H).

ASSOCIATED CONTENT

* Supporting Information

■

sı

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/acs.jmedchem.0c00627.

Additional biological results and chemical character-

ization data. File containing the pKa calculation, addi-

tional in vitro ADME data for 8a−d, 10a, 10b, 14a, 14b,

15a, 19a, and 19b (metabolic stability in dog hepatocytes

and human liver microsomes, plasma protein binding in

human and SD rat plasma, human cytochrome P450

inhibition, and/or metabolic proﬁling in human hep-

1925

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

ABBREVIATIONS

atocytes), additional data from the in vivo HFD study for

10b, 15a, and 19a (muscle glycogen content, glucose and

insulin from the OGTT, weight of selected tissues, and

body composition), results from the functional observa-

tion battery test for compound 10b, and chemical

characterization data for 4, 6, 7, 8a−j, 9, 10a, 10b, 14a,

14b, 15a, 15b, 19a, and 19b (¹H NMR, ¹³C NMR, and/or

LC-MS) (PDF)

■

ALT, alanine transaminase; ANCOVA, analysis of covariance;

AST, aspartate aminotransferase; B/P, brain-to-plasma; BW,

body weight; DFT, density functional theory; DIO, diet-induced

obesity; DMSO, dimethyl sulfoxide; FI, food intake; FOB,

functional observation battery; GLM, general linear model;

GPCRs, G protein-coupled receptors; HFD, high-fat diet; HLM,

human liver microsomes; HPMC, hydroxypropyl methylcellu-

lose; HRMS, high-resolution mass spectrometry; NAFLD, non-

alcoholic fatty liver disease; NASH, non-alcoholic Steatohepa-

titis; OGTT, oral glucose tolerance test; PEG400, polyethylene

glycol 400; PK, pharmacokinetics; SD, Sprague Dawley; SEM,

standard error of the mean; SGLT2, sodium-glucose cotrans-

porter-2

Molecular formula strings and in vitro ADME (logD, pK_a,

Caco-2 bidirectional permeability, kinetic solubility, and

metabolic stability in human, rat, and mouse hepatocytes)

and MT1/MT2 aﬃnity and potency data (CSV)

Binding poses for 10b (ACH-000143), 15a, and

reference compounds (ZIP)

REFERENCES

■

(1) Stenvers, D. J.; Scheer, F. A. J. L.; Schrauwen, P.; la Fleur, S. E.;

Kalsbeek, A. Circadian clocks and insulin resistance. Nat. Rev.

Endocrinol. 2019, 15, 75−89.

AUTHOR INFORMATION

■

Corresponding Author

̂

Hatylas Azevedo − Aché Laboratórios Farmaceuticos,

Guarulhos, Sao Paulo 07034-904, Brazil; orcid.org/0000-

0001-9233-0696; Email: hatylas.azevedo@ache.com.br

(2) Engin, A. Circadian rhythms in diet-induced obesity. Adv. Exp.

Med. Biol. 2017, 960, 19−52.

̃

(3) Shi, S. Q.; Ansari, T. S.; McGuinness, O. P.; Wasserman, D. H.;

Johnson, C. H. Circadian disruption leads to insulin resistance and

obesity. Curr. Biol. 2013, 23, 372−381.

Authors

Marcos Antonio Ferreira, Jr. − Aché Laboratórios

(4) Barclay, J. L.; Shostak, A.; Leliavski, A.; Tsang, A. H.; Jöhren, O.;

Muller-Fielitz, H.; Landgraf, D.; Naujokat, N.; van der Horst, G. T.;

̈

̂

Farmaceuticos, Guarulhos, Sao

Alessandra Mascarello − Aché Laboratórios Farmaceuticos,

Guarulhos, Sao Paulo 07034-904, Brazil

Natanael Dante Segretti − Aché Laboratórios Farmaceuticos,

Guarulhos, Sao Paulo 07034-904, Brazil

Elisa Russo − Zirkon Ind. Com de Insumos Químicos, Itapira,

Sao Paulo 13977-105, Brazil

Valter Russo − Zirkon Ind. Com de Insumos Químicos, Itapira,

Sao Paulo 13977-105, Brazil

Cristiano Ruch Werneck Guimara

̃

Paulo 07034-904, Brazil

Oster, H. High-fat diet-induced hyperinsulinemia and tissue-specific

insulin resistance in Cry-deficient mice. Am. J. Physiol. Endocrinol.

Metab. 2013, 304, E1053−E1063.

̂

̃

̂

(5) Oishi, K.; Ohkura, N.; Wakabayashi, M.; Shirai, H.; Sato, K.;

Matsuda, J.; Atsumi, G.; Ishida, N. CLOCK is involved in obesity-

induced disordered fibrinolysis in ob/ob mice by regulating PAI-1 gene

expression. J. Thromb. Haemost. 2006, 4, 1774−1780.

̃

(6) Chaix, A.; Lin, T.; Le, H. D.; Chang, M. W.; Panda, S. Time-

restricted feeding prevents obesity and metabolic syndrome in mice

lacking a circadian clock. Cell Metab. 2019, 29, 303−319.e4.

(7) Riestra, P.; Gebreab, S. Y.; Xu, R.; Khan, R. J.; Gaye, A.; Correa, A.;

Min, N.; Sims, M.; Davis, S. K. Circadian CLOCK gene polymorphisms

in relation to sleep patterns and obesity in African Americans: findings

from the Jackson heart study. BMC Genet. 2017, 18, 58.

̃

̂

Farmaceuticos, Guarulhos, Sao

̃

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.jmedchem.0c00627

(8) Valladares, M.; Obregón, A. M.; Chaput, J. P. Association between

genetic variants of the clock gene and obesity and sleep duration. J.

Physiol. Biochem. 2015, 71, 855−860.

(9) Karlsson, B.; Knutsson, A.; Lindahl, B. Is there an association

between shift work and having a metabolic syndrome? Results from a

population based study of 27,485 people. Occup. Environ. Med. 2001,

58, 747−752.

Author Contributions

^#M.F., H.A., and A.M. contributed equally to this work as ﬁrst

co-authors.

Author Contributions

E.R. and V.R. synthesized the compounds. M.F., H.A., A.M., and

C.R.W.G. designed the biological experiments and analyzed the

results. N.S. performed the docking studies. M.F., H.A., A.M.,

N.S., and C.R.W.G. drafted the manuscript. All authors reviewed

the ﬁnal version of the manuscript.

(10) Nikpour, M.; Tirgar, A.; Hajiahmadi, M.; Hosseini, A.; Heidari,

B.; Ghaffari, F.; Ebadi, A.; Nasiri, F.; Firouzbakht, M. Shift work and

metabolic syndrome: A multi-center cross-sectional study on females of

reproductive age. Biomed. Rep. 2019, 10, 311−317.

(11) Valenzuela, F. J.; Vera, J.; Venegas, C.; Mun

Munoz, K.; Lagunas, C. Evidences of polymorphism associated with

̃

oz, S.; Oyarce, S.;

Funding

̃

́

This study was fully funded by Ache Laboratorios Farm-

circadian system and risk of pathologies: a review of the literature. Int. J.

̂

Endocrinol. 2016, 2016, 2746909.

(12) Corella, D.; Asensio, E. M.; Coltell, O.; Sorlí, J. V.; Estruch, R.;

̃

Martínez-González, M. A.; Salas-Salvadó, J.; Castaner, O.; Arós, F.;

Lapetra, J.; Serra-Majem, L.; Gómez-Gracia, E.; Ortega-Azorín, C.; Fiol,

M.; Espino, J. D.; Díaz-López, A.; Fitó, M.; Ros, E.; Ordovás, J. M.

CLOCK gene variation is associated with incidence of type-2 diabetes

and cardiovascular diseases in type-2 diabetic subjects: dietary

modulation in the PREDIMED randomized trial. Cardiovasc. Diabetol.

2016, 15, 4.

aceuticos.

Notes

The authors declare the following competing ﬁnancial

interest(s): The authors declare they were employed by Ache

Laboratrios Farmacuticos or Zirkon during the performance of

́

this work. All studies were funded by Ache Laboratorios

̂

Farmaceuticos.

(13) Ferreira, M. V. N.; Goumidi, L.; Carvalho, H. B.; De Moraes, A.

C. F.; Santaliestra-Pasías, A. M.; Kafatos, A.; Molnar, D.; Lambrinou, C.

P.; De Henauw, S.; Gutierrez, A.; Censi, L.; Marcos, A.; Widhalm, K.;

Gottrand, F.; Gonzalez-Gross, M.; Meirhaeghe, A.; Moreno, L. A.

ACKNOWLEDGMENTS

■

We appreciate the analysis and the execution of the HFD

experiments by RenaSci.

1926

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

HELENA Study group. Associations between REV-ERB α , sleep

duration and body mass index in European adolescents. Sleep Med.

2018, 46, 56−60.

(30) Jockers, R.; Delagrange, P.; Dubocovich, M. L.; Markus, R. P.;

Renault, N.; Tosini, G.; Cecon, E.; Zlotos, D. P. Update on melatonin

receptors: IUPHAR review 20. Br. J. Pharmacol. 2016, 173, 2702−2725.

(31) Depreux, P.; Lesieur, D.; Mansour, H. A.; Morgan, P.; Howell, H.

E.; Renard, P.; Caignard, D. H.; Pfeiffer, B.; Delagrange, P.; Guardiola-

Lemaître, B.; Andrieux, J. Synthesis and structure-activity relationships

of novel naphthalenic and bioisosteric related amidic derivatives as

melatonin receptor ligands. J. Med. Chem. 1994, 37, 3231−3239.

(32) Zlotos, D. P.; Jockers, R.; Cecon, E.; Rivara, S.; Witt-Enderby, P.

A. MT1 and MT2 melatonin receptors: ligands, models, oligomers, and

therapeutic potential. J. Med. Chem. 2014, 57, 3161−3185.

(33) Koike, T.; Takai, T.; Hoashi, Y.; Nakayama, M.; Kosugi, Y.;

Nakashima, M.; Yoshikubo, S.; Hirai, K.; Uchikawa, O. Synthesis of a

novel series of tricyclic dihydrofuran derivatives: discovery of 8,9-

dihydrofuro[3,2-c]pyrazolo[1,5-a]pyridines as melatonin receptor

(MT1/MT2) ligands. J. Med. Chem. 2011, 54, 4207−4218.

(34) Stauch, B.; Johansson, L. C.; McCorvy, J. D.; Patel, N.; Han, G.

W.; Huang, X. P.; Gati, C.; Batyuk, A.; Slocum, S. T.; Ishchenko, A.;

Brehm, W.; White, T. A.; Michaelian, N.; Madsen, C.; Zhu, L.; Grant, T.

D.; Grandner, J. M.; Shiriaeva, A.; Olsen, R. H. J.; Tribo, A. R.; Yous, S.;

Stevens, R. C.; Weierstall, U.; Katritch, V.; Roth, B. L.; Liu, W.;

Cherezov, V. Structural basis of ligand recognition at the human MT1

melatonin receptor. Nature 2019, 569, 284−288.

(35) Johansson, L. C.; Stauch, B.; McCorvy, J. D.; Han, G. W.; Patel,

N.; Huang, X. P.; Batyuk, A.; Gati, C.; Slocum, S. T.; Li, C.; Grandner, J.

M.; Hao, S.; Olsen, R. H. J.; Tribo, A. R.; Zaare, S.; Zhu, L.; Zatsepin, N.

A.; Weierstall, U.; Yous, S.; Stevens, R. C.; Liu, W.; Roth, B. L.; Katritch,

V.; Cherezov, V. XFEL structures of the human MT2 melatonin

receptor reveal the basis of subtype selectivity. Nature 2019, 569, 289−

292.

(36) Avdeef, A. Physicochemical profiling (solubility, permeability

and charge state). Curr. Top. Med. Chem. 2001, 1, 277−351.

(37) Bochevarov, A. D.; Harder, E.; Hughes, T. F.; Greenwood, J. R.;

Braden, D. A.; Philipp, D. M.; Rinaldo, D.; Halls, M. D.; Zhang, J.;

Friesner, R. A. Jaguar: a high-performance quantum chemistry software

program with strengths in life and materials sciences. Int. J. Quantum

Chem. 2013, 113, 2110−2142.

(38) Depreux, P.; Fourmaintraux, E.; Lesieur, D.; Renard, P.

Regioselective synthesis of benzimidazolic bioisosteres of melatonin.

Synth. Commun. 1994, 24, 2123−2132.

(14) Cipolla-Neto, J.; Amaral, F. G.; Afeche, S. C.; Tan, D. X.; Reiter,

R. J. Melatonin, energy metabolism, and obesity: a review. J. Pineal Res.

2014, 56, 371−381.

(15) Bouatia-Naji, N.; Bonnefond, A.; Cavalcanti-Proenca̧ , C.; Sparsø,

T.; Holmkvist, J.; Marchand, M.; Delplanque, J.; Lobbens, S.;

̀

Rocheleau, G.; Durand, E.; de Graeve, F.; Chevre, J. C.; Borch-

Johnsen, K.; Hartikainen, A. L.; Ruokonen, A.; Tichet, J.; Marre, M.;

Weill, J.; Heude, B.; Tauber, M.; Lemaire, K.; Schuit, F.; Elliott, P.;

Jørgensen, T.; Charpentier, G.; Hadjadj, S.; Cauchi, S.; Vaxillaire, M.;

Sladek, R.; Visvikis-Siest, S.; Balkau, B.; Lévy-Marchal, C.; Pattou, F.;

Meyre, D.; Blakemore, A. I.; Jarvelin, M. R.; Walley, A. J.; Hansen, T.;

Dina, C.; Pedersen, O.; Froguel, P. A variant near MTNR1B is

associated with increased fasting plasma glucose levels and type 2

diabetes risk. Nat. Genet. 2009, 41, 89−94.

(16) Tuomi, T.; Nagorny, C. L. F.; Singh, P.; Bennet, H.; Yu, Q.;

̈

Alenkvist, I.; Isomaa, B.; Ostman, B.; Söderström, J.; Pesonen, A. K.;

Martikainen, S.; Räikkönen, K.; Forsén, T.; Hakaste, L.; Almgren, P.;

Storm, P.; Asplund, O.; Shcherbina, L.; Fex, M.; Fadista, J.; Tengholm,

A.; Wierup, N.; Groop, L.; Mulder, H. Increased melatonin signaling is a

risk factor for type 2 diabetes. Cell Metab. 2016, 23, 1067−1077.

(17) Sun, H.; Wang, X.; Chen, J.; Song, K.; Gusdon, A. M.; Li, L.; Bu,

L.; Qu, S. Melatonin improves non-alcoholic fatty liver disease via

MAPK-JNK/P38 signaling in high-fat-diet-induced obese mice. Lipids

Health Dis. 2016, 15, 202.

(18) Tahan, V.; Atug, O.; Akin, H.; Eren, F.; Tahan, G.; Tarcin, O.;

Uzun, H.; Ozdogan, O.; Tarcin, O.; Imeryuz, N.; Ozguner, F.; Celikel,

C.; Avsar, E.; Tozun, N. Melatonin ameliorates methionine- and

choline-deficient diet-induced nonalcoholic steatohepatitis in rats. J.

Pineal Res. 2009, 46, 401−407.

(19) Thomas, A. P.; Hoang, J.; Vongbunyong, K.; Nguyen, A.;

Rakshit, K.; Matveyenko, A. V. Administration of melatonin and

metformin prevents deleterious effects of circadian disruption and

obesity in male rats. Endocrinology 2016, 157, 4720−4731.

(20) Tordjman, S.; Chokron, S.; Delorme, R.; Charrier, A.; Bellissant,

E.; Jaafari, N.; Fougerou, C. Melatonin: pharmacology, functions and

therapeutic benefits. Curr. Neuropharmacol. 2017, 15, 434−443.

(21) Emet, M.; Ozcan, H.; Ozel, L.; Yayla, M.; Halici, Z.;

Hacimuftuoglu, A. A review of melatonin, its receptors and drugs.

Eurasian J. Med. 2016, 48, 135−141.

(39) Cacatian, S.; Claremon, D. A.; Dillard, L. W.; Dong, C.; Fan, Y.;

Houston, T. L.; Jia, L.; Lotesta, S. D.; Marcus, A.; Morales-Ramos, A.;

Parent, S. D.; Singh, S. D.; Venkatraman, S.; Yuan, J.; Zheng, Y.;

Zhuang, L. Inhibitors of the Menin-MLL Interaction. 2017,

WO2017214367 A1.

(22) Liu, J.; Clough, S. J.; Hutchinson, A. J.; Adamah-Biassi, E. B.;

Popovska-Gorevski, M.; Dubocovich, M. L. MT1 and MT2 melatonin

receptors: a therapeutic perspective. Annu. Rev. Pharmacol. Toxicol.

2016, 56, 361−383.

̌

́

(40) Cotarca, L.; Delogu, P.; Nardelli, A.; Sunjic, V. Bis-

(trichloromethyl) carbonate in organic synthesis. Synthesis 1996, 5,

553−576.

(23) Wojcik, M.; Krawczyk, M.; Wojcik, P.; Cypryk, K.; Wozniak, L. A.

Melatonin as a pleiotropic molecule with therapeutic potential for type

2 diabetes and cancer. Curr. Med. Chem. 2017, 24, 3829−3850.

(41) Mulchahey, J. J.; Goldwater, D. R.; Zemlan, F. P. A single blind,

placebo controlled, across groups dose escalation study of the safety,

tolerability, pharmacokinetics and pharmacodynamics of the melatonin

analog beta-methyl-6-chloromelatonin. Life Sci. 2004, 75, 1843−1856.

(42) Zhu, J.; Hu, Y.; Ho, M. K. C.; Wong, Y. H. 3-

Methoxylphenylpropyl amides as novel receptor subtype-selective

melatoninergic ligands: characterization of physicochemical and

pharmacokinetic properties. Xenobiotica 2011, 41, 35−45.

(43) Yeleswaram, K.; McLaughlin, L. G.; Knipe, J. O.; Schabdach, D.

Pharmacokinetics and oral bioavailability of exogenous melatonin in

preclinical animal models and clinical implications. J. Pineal Res. 1997,

22, 45−51.

(24) Acuna-Castroviejo, D.; Escames, G.; Venegas, C.; Díaz-Casado,

̃

M. E.; Lima-Cabello, E.; López, L. C.; Rosales-Corral, S.; Tan, D. X.;

Reiter, R. J. Extrapineal melatonin: sources, regulation, and potential

functions. Cell. Mol. Life Sci. 2014, 71, 2997−3025.

(25) Liu, L.; Labani, N.; Cecon, E.; Jockers, R. Melatonin target

proteins: too many or not enough? Front. Endocrinol. 2019, 10, 791.

(26) Gbahou, F.; Cecon, E.; Viault, G.; Gerbier, R.; Jean-Alphonse, F.;

Karamitri, A.; Guillaumet, G.; Delagrange, P.; Friedlander, R. M.;

Vilardaga, J. P.; Suzenet, F.; Jockers, R. Design and validation of the first

cell-impermeant melatonin receptor agonist. Br. J. Pharmacol. 2017,

174, 2409−2421.

(27) Vitte, P. A.; Harthe, C.; Lestage, P.; Claustrat, B.; Bobillier, P.

Plasma, cerebrospinal fluid, and brain distribution of 14C-melatonin in

rat: a biochemical and autoradiographic study. J. Pineal Res. 1988, 5,

437−453.

(44) Flo, A.; Cambras, T.; Díez-Noguera, A.; Calpena, A. Melatonin

pharmacokinetics after transdermal administration changes according

to the time of the day. Eur. J. Pharm. Sci. 2017, 96, 164−170.

(45) Bowes, J.; Brown, A. J.; Hamon, J.; Jarolimek, W.; Sridhar, A.;

Waldron, G.; Whitebread, S. Reducing safety-related drug attrition: the

use of in vitro pharmacological profiling. Nat. Rev. Drug. Discov. 2012,

11, 909−922.

(28) Boutin, J. A.; Witt-Enderby, P. A.; Sotriffer, C.; Zlotos, D. P.

Melatonin receptor ligands: A pharmaco-chemical perspective. J. Pineal

Res. 2020, 69, No. e12672.

(29) Zlotos, D. P. Recent advances in melatonin receptor ligands.

Arch. Pharm. (Weinheim). 2005, 338, 229−247.

(46) Millar, P.; Pathak, N.; Parthsarathy, V.; Bjourson, A. J.; O’Kane,

M.; Pathak, V.; Moffett, R. C.; Flatt, P. R.; Gault, V. A. Metabolic and

1927

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

neuroprotective effects of dapagliflozin and liraglutide in diabetic mice.

J. Endocrinol. 2017, 234, 255−267.

adenosine 3 ′,5 ′-monophosphate-dependent signal transduction cas-

cade in Chinese hamster ovary cells expressing the human mt1

melatonin receptor. Endocrinology 1998, 139, 3064−3071.

(47) Devenny, J. J.; Godonis, H. E.; Harvey, S. J.; Rooney, S.; Cullen,

M. J.; Pelleymounter, M. A. Weight loss induced by chronic

dapagliflozin treatment is attenuated by compensatory hyperphagia in

diet-induced obese (DIO) rats. Obesity 2012, 20, 1645−1652.

(48) Bechmann, L. P.; Hannivoort, R. A.; Gerken, G.; Hotamisligil, G.

S.; Trauner, M.; Canbay, A. The interaction of hepatic lipid and glucose

metabolism in liver diseases. J. Hepatol. 2012, 56, 952−964.

(49) Scheen, A. J. Beneficial effects of SGLT2 inhibitors on fatty liver

in type 2 diabetes: A common comorbidity associated with severe

complications. Diabetes Metab. 2019, 45, 213−223.

(65) Owino, S.; Sánchez-Bretano, A.; Tchio, C.; Cecon, E.; Karamitri,

̃

A.; Dam, J.; Jockers, R.; Piccione, G.; Noh, H. L.; Kim, T.; Kim, J. K.;

Baba, K.; Tosini, G. Nocturnal activation of melatonin receptor type 1

signaling modulates diurnal insulin sensitivity via regulation of PI3K

activity. J. Pineal Res. 2018, 64, e12462.

(66) Beresford, I. J.; Browning, C.; Starkey, S. J.; Brown, J.; Foord, S.

M.; Coughlan, J.; North, P. C.; Dubocovich, M. L.; Hagan, R. M.

GR196429: a nonindolic agonist at high-affinity melatonin receptors. J.

Pharmacol. Exp. Ther. 1998, 285, 1239−1245.

(67) Browning, C.; Beresford, I.; Fraser, N.; Giles, H. Pharmacological

characterization of human recombinant melatonin mt1 and MT2

receptors. Br. J. Pharmacol. 2000, 129, 877−886.

(68) Witt-Enderby, P. A.; Dubocovich, M. L. Characterization and

regulation of the human ML1A melatonin receptor stably expressed in

Chinese hamster ovary cells. Mol. Pharmacol. 1996, 50, 166−174.

(69) Sastry, G. M.; Adzhigirey, M.; Day, T.; Annabhimoju, R.;

Sherman, W. Protein and ligand preparation: Parameters, protocols,

and influence on virtual screening enrichments. J. Comput.-Aided Mol.

Des. 2013, 27, 221−234.

(70) Shelley, J. C.; Cholleti, A.; Frye, L. L.; Greenwood, J. R.; Timlin,

M. R.; Uchimaya, M. Epik: a software program for pK(a) prediction and

protonation state generation for drug-like molecules. J. Comput.-Aided

Mol. Des. 2007, 21, 681−691.

(71) Li, H.; Robertson, A. D.; Jensen, J. H. Very fast empirical

prediction and rationalization of protein pKa values. Proteins 2005, 61,

704−721.

(72) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye,

L. L.; Pollard, W. T.; Banks, J. L. Glide: a new approach for rapid,

accurate docking and scoring. 2. Enrichment factors in database

screening. J. Med. Chem. 2004, 47, 1750−1759.

(73) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.

Experimental and computational approaches to estimate solubility and

permeability in drug discovery and development settings. Adv. Drug

Delivery Rev. 2001, 46, 3−26.

(74) Hidalgo, I. J.; Raub, T. J.; Borchardt, R. T. Characterization of the

human colon carcinoma cell line (Caco-2) as a model system for

intestinal epithelial permeability. Gastroenterology 1989, 96, 736−749.

(75) Obach, R. S.; Baxter, J. G.; Liston, T. E.; Silber, B. M.; Jones, B.

C.; MacIntyre, F.; Rance, D. J.; Wastall, P. The prediction of human

pharmacokinetic parameters from preclinical and in vitro metabolism

data. J. Pharmacol. Exp. Ther. 1997, 283, 46−58.

(50) Tølbøl, K. S.; Kristiansen, M. N.; Hansen, H. H.; Veidal, S. S.;

Rigbolt, K. T.; Gillum, M. P.; Jelsing, J.; Vrang, N.; Feigh, M. Metabolic

and hepatic effects of liraglutide, obeticholic acid and elafibranor in

diet-induced obese mouse models of biopsy-confirmed nonalcoholic

steatohepatitis. World J. Gastroenterol. 2018, 24, 179−194.

(51) Yokono, M.; Takasu, T.; Hayashizaki, Y.; Mitsuoka, K.; Kihara,

R.; Muramatsu, Y.; Miyoshi, S.; Tahara, A.; Kurosaki, E.; Li, Q.;

Tomiyama, H.; Sasamata, M.; Shibasaki, M.; Uchiyama, Y. SGLT2

selective inhibitor ipragliflozin reduces body fat mass by increasing fatty

acid oxidation in high-fat diet-induced obese rats. Eur. J. Pharmacol.

2014, 727, 66−74.

(52) Redfern, W. S.; Strang, I.; Storey, S.; Heys, C.; Barnard, C.;

Lawton, K.; Hammond, T. G.; Valentin, J. P. Spectrum of effects

detected in the rat functional observational battery following oral

administration of non-CNS targeted compounds. J. Pharmacol. Toxicol.

Methods 2005, 52, 77−82.

(53) Redfern, W. S.; Dymond, A.; Strang, I.; Storey, S.; Grant, C.;

Marks, L.; Barnard, C.; Heys, C.; Moyser, K.; Greenwood, K.; Cobey,

D.; Moore, N.; Karp, N. A.; Prior, H. The functional observational

battery and modified Irwin test as global neurobehavioral assessments

in the rat: Pharmacological validation data and a comparison of

methods. J. Pharmacol. Toxicol. Methods 2019, 98, 106591.

(54) Harpsøe, N. G.; Andersen, L. P. H.; Gögenur, I.; Rosenberg, J.

Clinical pharmacokinetics of melatonin: a systematic review. Eur. J.

Clin. Pharmacol. 2015, 71, 901−909.

(55) Pereira, M. J.; Eriksson, J. W. Emerging role of SGLT-2 inhibitors

for the treatment of obesity. Drugs 2019, 79, 219−230.

(56) Pan, M.; Song, Y. L.; Xu, J. M.; Gan, H. Z. Melatonin ameliorates

nonalcoholic fatty liver induced by high-fat diet in rats. J. Pineal Res.

2006, 41, 79−84.

(57) Hatzis, G.; Ziakas, P.; Kavantzas, N.; Triantafyllou, A.; Sigalas, P.;

Andreadou, I.; Ioannidis, K.; Chatzis, S.; Filis, K.; Papalampros, A.;

Sigala, F. Melatonin attenuates high fat diet-induced fatty liver disease

in rats. World J. Hepatol. 2013, 5, 160−169.

(58) Karamitri, A.; Jockers, R. Melatonin in type 2 diabetes mellitus

and obesity. Nat. Rev. Endocrinol. 2019, 15, 105−125.

(59) Gonciarz, M.; Gonciarz, Z.; Bielanski, W.; Mularczyk, A.;

Konturek, P. C.; Brzozowski, T.; Konturek, S. J. The effects of long-term

melatonin treatment on plasma liver enzymes levels and plasma

concentrations of lipids and melatonin in patients with nonalcoholic

steatohepatitis: a pilot study. J. Physiol. Pharmacol. 2012, 63, 35−40.

(60) Besag, F. M. C.; Vasey, M. J.; Lao, K. S. J.; Wong, I. C. K. Adverse

events associated with melatonin for the treatment of primary or

secondary sleep disorders: a systematic review. CNS Drugs 2019, 33,

1167−1186.

(61) van den Berg, M. P.; Merkus, P.; Romeijn, S. G.; Verhoef, J. C.;

Merkus, F. W. H. M. Uptake of melatonin into the cerebrospinal fluid

after nasal and intravenous delivery: studies in rats and comparison with

a human study. Pharm. Res. 2004, 21, 799−802.

(62) Contreras-Alcantara, S.; Baba, K.; Tosini, G. Removal of

melatonin receptor type 1 induces insulin resistance in the mouse.

Obesity (Silver Spring). 2010, 18, 1861−1863.

(63) Owino, S.; Buonfiglio, D. D. C.; Tchio, C.; Tosini, G. Melatonin

signaling a key regulator of glucose homeostasis and energy

metabolism. Front. Endocrinol. 2019, 10, 488.

(76) Ono, S.; Hatanaka, T.; Hotta, H.; Satoh, T.; Gonzalez, F. J.;

Tsutsui, M. Specificity of substrate and inhibitor probes for cytochrome

P450s: evaluation of in vitro metabolism using cDNA-expressed human

P450s and human liver microsomes. Xenobiotica 2000, 26, 681−693.

(77) Miller, V. P.; Stresser, D. M.; Blanchard, A. P.; Turner, S.; Crespi,

C. L. Fluorometric high-throughput screening for inhibitors of

cytochrome P450. Ann. N. Y. Acad. Sci. 2000, 919, 26−32.

(78) Banker, M. J.; Clark, T. H.; Williams, J. A. Development and

validation of a 96-well equilibrium dialysis apparatus for measuring

plasma protein binding. J. Pharm. Sci. 2003, 92, 967−974.

(79) Wenlock, M. C.; Potter, T.; Barton, P.; Austin, R. P. A method for

measuring the lipophilicity of compounds in mixtures of 10. J. Biomol.

Screen. 2011, 16, 348−355.

(80) Tam, K. Y.; Takács-Novák, K. Multi-wavelength spectrophoto-

metric determination of acid dissociation constants: a validation study.

Anal. Chim. Acta 2001, 434, 157−167.

(81) Takács-Novák, K.; Box, K. J.; Avdeef, A. Potentiometric pK_a

determination of water-insoluble compounds: validation study in

methanol:water mixtures. Int. J. Pharm. 1997, 151, 235−248.

(82) Prunet-Marcassus, B.; DesBazeille, M.; Bros, A.; Louche, K.;

DelaGrange, P.; Renard, P.; Casteilla, P.; Penicaud, L. Melatonin

reduces body weight gain in Sprague Dawley rats with diet-induced

obesity. Endocrinology 2003, 144, 5347−5352.

(64) Witt-Enderby, P. A.; Masana, M. I.; Dubocovich, M. L.

Physiological exposure to melatonin supersensitizes the cyclic

(83) van Dijk, T. H.; Laskewitz, A. J.; Grefhorstm, A.; Boer, T. S.;

Bloks, V. W.; Kuipers, F.; Groen, A. K.; Reijngoud, D. J. A novel

1928

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Journal of Medicinal Chemistry

pubs.acs.org/jmc

Article

approach to monitor glucose metabolism using stable isotopically

labelled glucose in longitudinal studies in mice. Lab. Anim. 2013, 47,

79−88.

(84) Kleiner, D. E.; Brunt, E. M.; Van Natta, M.; Behling, C.; Contos,

M. J.; Cummings, O. W.; Ferrell, L. D.; Liu, Y. C.; Torbenson, M. S.;

Unalp-Arida, A.; Yeh, M.; McCullough, A. J.; Sanyal, A. J. Nonalcoholic

steatohepatitis clinical research network. design and validation of a

histological scoring system for nonalcoholic fatty liver disease.

Hepatology 2005, 41, 1313−1321.

(85) Nishida, T.; Tsuneyama, K.; Fujimoto, M.; Nomoto, K.; Hayashi,

S.; Miwa, S.; Nakajima, T.; Nakanishi, Y.; Sasaki, Y.; Suzuki, W.; Iizuka,

S.; Nagata, M.; Shimada, T.; Aburada, M.; Shimada, Y.; Imura, J.

Spontaneous onset of nonalcoholic steatohepatitis and hepatocellular

carcinoma in a mouse model of metabolic syndrome. Lab. Invest. 2013,

93, 230−241.

(86) Irwin, S. Comprehensive observational assessment: Ia. A

systematic, quantitative procedure for assessing the behavioral and

physiologic state of the mouse. Psychopharmacologia 1968, 13, 222−

257.

(87) Maron, D. M.; Ames, B. N. Revised methods for the Salmonella

mutagenicity test. Mutat. Res. 1983, 113, 173−215.

NOTE ADDED AFTER ISSUE PUBLICATION

This article was initially published with an incorrect copyright

statement and was corrected on or around May 5, 2021.

■

1929

https://doi.org/10.1021/acs.jmedchem.0c00627

J. Med. Chem. 2021, 64, 1904−1929

Article Doi

Discovery of ACH-000143: A Novel Potent and Peripherally Preferred Melatonin Receptor Agonist that Reduces Liver Triglycerides and Steatosis in Diet-Induced Obese Rats

DOI: 10.1021/acs.jmedchem.0c00627

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.0c00627

Full text of DOI:10.1021/acs.jmedchem.0c00627

Products guided by the article

R&D Labs maybe for 13556-31-5

Relevant to this article

Hot Product