Discovery of N-Arylsulfonyl-Indole-2-Carboxamide Derivatives as Potent, Selective, and Orally Bioavailable Fructose-1,6-Bisphosphatase Inhibitors - Design, Synthesis, in Vivo Glucose Lowering Effects, and X-ray Crystal Complex Analysis

Article abstract of DOI:10.1021/acs.jmedchem.0c00726

Liver fructose-1,6-bisphosphatase (FBPase) is a key enzyme in the gluconeogenesis pathway. Inhibiting FBPase activity represents a potential treatment for type 2 diabetes mellitus. A series of novel N-arylsulfonyl-4-arylamino-indole-2-carboxamide derivatives have been disclosed as FBPase inhibitors. Through extensive structure-activity relationship investigations, a promising candidate molecule Cpd118 [sodium (7-chloro-4-((3-methoxyphenyl)amino)-1-methyl-1H-indole-2-carbonyl] [(4-methoxyphenyl)sulfonyl)amide] has been identified with high inhibitory activity against human liver FBPase (IC50, 0.029 ± 0.006 μM) and high selectivity relative to the other six AMP-binding enzymes. Importantly, Cpd118 produced significant glucose-lowering effects on both type 2 diabetic KKAy mice and ZDF rats as demonstrated by substantial reductions in the fasting and postprandial blood glucose levels, as well as the HbA1c level. Furthermore, Cpd118 elicited a favorable pharmacokinetic profile with an oral bioavailability of 99.1%. Moreover, the X-ray crystal structure of the Cpd118-FBPase complex was resolved, which revealed a unique binding mode and provided a structural basis for its high potency and selectivity.

Full text of DOI:10.1021/acs.jmedchem.0c00726

Subscriber access provided by Gothenburg University Library
Article
Discovery of N-arylsulfonyl-indole-2-carboxamide Derivatives
as Potent, Selective, and Orally Bioavailable Fructose-1,6-
Bisphosphatase Inhibitors# Design, Synthesis, In Vivo
Glucose Lowering Effects, and X-ray Crystal Complex Analysis
Jie Zhou, Jianbo Bie, Xiaoyu Wang, Quan Liu, Rongcui Li, Hualong Chen, Jinping
Hu, Hui Cao, Wenming Ji, Yan Li, Shuainan Liu, Zhu-fang Shen, and Bailing Xu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00726 • Publication Date (Web): 21 Aug 2020
Downloaded from pubs.acs.org on August 22, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of N-arylsulfonyl-indole-2-carboxamide Derivatives as Potent,
Selective, and Orally Bioavailable Fructose-1,6-Bisphosphatase Inhibitors―
Design, Synthesis, In Vivo Glucose Lowering Effects, and X-ray Crystal
Complex Analysis
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
,‖
†,‖
†,‖
‡
‡
†
Jie Zhou , Jianbo Bie , Xiaoyu Wang , Quan Liu , Rongcui Li , Hualong Chen ,
§
‡
‡
§
‡,*
‡,*
Jinping Hu , Hui Cao , Wenming Ji , Yan Li , Shuainan Liu , Zhufang Shen ,
Bailing Xu†,*
†Beijing Key Laboratory of Active Substances Discovery and Druggability Evaluation,
‡
§
department of pharmacology, and Beijing Key Laboratory of Non-Clinical Drug Metabolism
and PK/PD Study, Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking
Union Medical College, Beijing 100050, China
ABSTRACT: Liver fructose-1,6-bisphosphatase (FBPase) is a key enzyme in the
gluconeogenesis pathway. Inhibiting FBPase activity represents a potential treatment
for
type
2
diabetes
mellitus.
A
series
of
novel
N-arylsulfonyl-4-arylamino-indole-2-carboxamide derivatives have been disclosed as
FBPase inhibitors. Through extensive structure-activity relationship investigations,
a
promising
candidate
molecule
Cpd118
(sodium
(7-chloro-4-((3-methoxyphenyl)amino)-1-methyl-1H-indole-2-carbonyl)((4-methoxyp
henyl)sulfonyl)amide) has been identified with high inhibitory activity against human
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
liver FBPase (IC , 0.029±0.006 μM) and high selectivity relative to the other six
5
0
AMP-binding enzymes. Importantly, Cpd118 produced significant glucose-lowering
effects on both type 2 diabetic KKAy mice and ZDF rats as demonstrated by
substantial reductions in the fasting and postprandial blood glucose levels, as well as
the HbA1c level. Furthermore, Cpd118 elicited a favorable pharmacokinetic profile
with the oral bioavailability of 99.1%. Moreover, the X-ray crystal structure of the
Cpd118-FBPase complex was resolved, which revealed a unique binding mode and
provided a structural basis for its high potency and selectivity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0

INTRODUCTION
Diabetes mellitus is one of the most prevalent chronic diseases worldwide. According
to the ninth edition report released by IDF in 2019, approximately 463 million adults
(about 9.3% of the population) are suffering from diabetes globally, and this figure is
1
predicted to increase up to 578 million (about 10.2%) by the year of 2030. Diabetes
is a metabolic disorder diagnosed by high blood glucose levels (hyperglycemia) in
both fasting and postprandial state of patients. The main clinical forms of diabetes
include type 1 diabetes mellitus (T1DM), type 2 diabetes mellitus (T2DM), as well as
gestational diabetes mellitus, with T2DM accounting for the majority (around 90%) of
diagnosed diabetes worldwide. It has been demonstrated that long-term
hyperglycemia has been more likely to lead to various debilitating microvascular and
2
3
4
macrovascular complications, such as retinopathy and cataract, nephropathy,
5
6
neuropathy and cardiovascular diseases. Therefore, the control of glucose level is
ACS Paragon Plus Environment

Page 3 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the primary choice for the treatment of diabetes and its associate complications.
The endogenous glucose production (EGP) in liver plays a key role in regulation of
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
blood glucose level and is positively correlated with fasting blood glucose level. In
the fasting state, glucose is solely produced by EGP processes including
glycogenolysis and gluconeogenesis. Inhibiting EGP represents a promising strategy
for glycemic control. However, no anti-diabetic drugs directly targeting on EGP
process are currently available.
Gluconeogenesis is a primary EGP process for liver to produce endogenous
glucose, which involves biochemical transformations of
a
number of
non-carbohydrate precursors such as glycerol, alanine and lactate into glucose with
the catalysis of several key enzymes including phosphoenopyruvate carboxykinase
(PEPCK), fructose-1,6-bisphosphatase (FBPase) and glucose-6-phosphatase
8
(
G-6-Pase). During long periods of fasting or starvation, gluconeogenesis contributes
to more than 50% glucose production, and this contribution is even more significant
in T2DM patients.9,10 Therefore, targeting on the gluconeogenesis pathway is an
appealing approach for the management of T2DM.
FBPase catalyzes the conversion of fructose-1,6-bisphosphate into
fructose-6-phosphate and inorganic phosphate. 11 It comprises two isoforms in
mammals, namely liver FBPase (encoded by FBP1 gene, referred to as FBPase or
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
FBPase1) and muscle FBPase (encoded by FBP2 gene, referred to as FBPase2). Liver
FBPase, mainly expressed in liver and kidney, is a rate-limiting enzyme in hepatic
gluconeogenesis pathway, and plays a key role in the regulation of glucose levels.
Muscle FBPase is widely expressed in mammal cells including muscle tissue, neurons
and liver, functioning as a multifaceted regulator in cells.12-14 Additionally, it has been
observed that liver FBPase is elevated in insulin-resistant and insulin-deficient animal
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
5
models of diabetes. Therefore, FBPase inhibitors have been intensively explored as
anti-diabetic agents for the direct regulation of EGP.
Human liver FBPase is a homotetramer with each monomer consisting of 337
amino acid residues. Three distinct binding sites exist within the tetramer of FBPase:
a fructose-1,6-biphosphate binding site and an AMP binding site within each
monomer, and a binding pocket situated at the interface between the upper dimer and
1
6
lower dimer. Among these binding sites, the AMP binding pocket has received more
attention because it is presumably a more druggable site compared with the other two
sites. The AMP binding allosteric site is 28Å away from the substrate site. Upon
binding the AMP to the allosteric site, the quaternary conformation of FBPase could
change from the active R state to the inactive T state, inhibiting the catalytic activity
of FBPase.13,17
Up to date, a number of FBPase inhibitors bearing distinct scaffolds have been
identified via either high-throughput screening of compound libraries or
ACS Paragon Plus Environment

Page 5 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
structure-guided design of AMP mimetics.18-31 Presently, only two phosphonic
acid-containing thiazole derivatives were advanced into clinical trials (Figure 1). 32
The clinical trial of the first candidate CS-917 was terminated in 2008 due to the
toxicity of its metabolite, and the second candidate MB07803 is now in phase II
clinical trial.33,34 Also, these two candidates bearing the phosphonic acid group were
developed as phosphonic diamide prodrugs to achieve the oral bioavailability.
Evidently, it is challengeing to develop potent and orally bioavailable FBPase
inhibitors without a phosphonic acid group.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The indole derivative MDL-29951 was found to be a potent FBPase inhibitor by a
high throughput screening in Pfizer.18 To search for drug-like FBPase inhibitors
without a phosphonate or phosphate group, MDL-29951 was chosen as our starting
point, and the investigations on structure-activity relationships (SARs) led to a new
hit compound A (Figure 2). 35 Taking advantage of N-acyl sulfonamide moiety as the
bioisosteric replacement of the carboxyl group, compound B (Figure 2) was
designed, and Cpd118 was discovered as a potent and orally bioavailable FBPase
inhibitor. Herein, the design, synthesis and extensive SAR investigations of
N-arylsulfonyl-indole-2-carboxamide derivatives were presented. The in vivo glucose
lowering efficacy in diabetic type 2 animal models and the pharmacokinetic
properties of Cpd118 were disclosed. Furthermore, the cocrystal structure of Cpd118
bound to FBPase was solved and a distinct binding mode was revealed.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EtOOC
OH
O
OH
HN
P
O
O
H2N
P
H N
H2N
P
N
O
2
N
O
N
O
OH
OH
NH
S
S
S
COOEt
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
MB05032
IC = 0.025 μΜ
CS917
MB07229
IC = 0.024 μΜ
5
0
50
CO2Et
HN
P
Cl
O
O
H2N
N
O
NH
CO2Et
O
N
S
O
O
Br
Cl
S
O Cl
Cl
S
N
O
S
O
N
N
NH
H
H
MB07803
IC = 3.4 μΜ
IC = 0.33 μΜ
50
5
0
Figure 1. Several reported FBPase inhibitors
OCH3
R9
O
OH
X
R10
R5
NH
Cl
O
O
O
R6
N
HN S
R1
O
O
N
OH
R7
N
OH
Cl
H
H
NO2
50
MDL-29951
A IC = 25.6 μM
B
Figure 2. The design and optimization of N-arylsulfonyl-indole-2-carboxamides as FBPase
inhibitors

RESULTS AND DISCUSSION
Molecular Design. The AMP binding allosteric site of FBPase consists of two
featured sub-pockets. One is an adenine binding hydrophobic pocket lined with
Val17, Ala24, Leu30, and Met177, and the other is a phosphate binding pocket where
the phosphate formed a hydrogen binding network with key amino acids including
36
Thr27, Gly28, Glu29, Lys112 and Tyr113 (Figure 3B-3D). The phosphate group
significantly contributes to the binding affinity; however, the inhibitors bearing a
ACS Paragon Plus Environment

Page 7 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
phosphate moiety confer a poor membrane permeability, which leads to low
bioavailability.37-39 Therefore, we focused on the exploration of inhibitors bearing
other acidic groups. As a result, indole-2-carboxylic acid derivatives were identified
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
5
as FBPase inhibitors. Among them, compound A was selected tentatively as the hit
structure for further optimization since its structure was quite distinct from known
inhibitors. At this point, we utilized the bioisosteric group N-acyl sulfonamide in
substitution for the carboxylic acid, constructing the N-sulfonyl-indole-2-carboxamide
derivatives (compound B). Delightfully, this bioisosteric replacement worked quite
effectively and resulted in the highly potent N-phenylsulfonyl-indole-2-carboxamide
derivative (compound 32), with inhibitory potency being improved by 182-fold than
that of hit compound A. Subsequently, the extensive structural modifications were
carried out around the N-arylsulfonyl-indole-2-carboxamide scaffold, furnishing the
candidate molecule Cpd118. Particularly, the X-ray cocrystal structure of Cpd118
bound to FBPase demonstrated that this class of inhibitors displayed a distinct binding
mode, which was presumably different from the hit compound A.
Chemistry. The chemical structures of target compounds (30-118) were shown in
Tables 1-7, and their preparations were illustrated in Schemes 1-3. The key
intermediate indole-2-carboxylic acid derivatives were synthesized by using the
Fischer indole synthesis and Hemetsberger indole synthesis, and their further
condensations with various aryl sulfonamides delivered the desired target compounds.
The preparation of 7-nitro substituted indole molecules involving Fischer indole
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
synthesis was outlined in Scheme 1. Upon treatment with NaNO /HCl, followed by
2
SnCl /HCl, 4-substituted-5-chloro-2-nitroanilines 1a-1d afforded hydrazine
2
hydrochlorides 2a-2d. The condensation between 2a-2d and ethyl pyruvate generated
hydrazone derivatives 3a-3d, which then transformed into the key indole
intermediates 4a-4d in 13-38% yields at 80 ℃ in PPA. Under the catalysis of
Pd (dba) /DavePhos, the coupling reaction of compounds 4a-4d with respectively
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
ArNH , aliphatic amines, benzylboronic acid pinacol ester, ArOH, or ArSH, yielded
2
5
a1-5a33 and 5b-5d (53-98%). The NaOH-catalyzed hydrolysis of 5a1-5a33, 5b-5d
produced 1H-indole-2-carboxylic acid derivatives (6a1-6a33, 6b-6d) in 37-99%
yields. The construction of target compounds (30–53, 55-78) was realized in
acceptable yields via the condensation of acid intermediates (6a1-6a33, 6b-6d) with
various sulfonamide compounds. Hydrolysis of compound 4a yielded 7a and
subsequently condensed with benzenesulfonamide, producing a 4-chloro-substituted
compound 54. Moreover, starting from compound 4a, N -substituted
1
indole-2-carboxylic acid derivatives 10a1-10a9 were obtained after three steps of
consecutive chemical transformations: alkylation with alkyl halides,
Buchwald-Hartwig coupling reaction and hydrolysis reaction. Finally, the
condensation between indole-2-carboxylic acid derivatives 10a1-10a9 and
4-methoxybenzenesulfonamide gave rise to target compounds 87-95 in 16-91%
yields.
ACS Paragon Plus Environment

Page 9 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Cl
Cl
Cl
Cl
R
R
R
R
a
R
O
b
EtO
NHN
c
d
HCl
.
O
N
OEt
NH2
NHNH
2
H
NO2
NO2
R
NO2
4a-4d
NO2
1a-1d
2a-2d
3a-3d
R4
R4
R4
R
O
O
O
1a-6a: R = H
1b-6b: R = F
1c-6c: R = Cl
1d-6d: R = Me
e
f
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
N
H
OEt
N
H
OH
N
HN
S
R8
H
NO2
NO2
NO2
5
a1-5a33, 5b-5d
6a1-6a33, 6b-6d
30-53, 55-78
Cl
Cl
Cl
O
O
OH
O
e
f
O
O
N
H
OEt
N
H
N
H
HN
54
R9
S
NO2
NO2
NO2
4a
7a
R9
Cl
Cl
HN
HN
O
O
g
d
O
e
O
N
H
OEt
N
OEt
N
OEt
N
OH
R1
R1
R1
NO2
NO2
NO2
9a1-9a9
NO2
4a
8a1-8a3
1
0a1-10a9
R9
HN
8a1, 9a1, 9a4-9a9, 10a1, 10a4-10a9: R1 = Me
O
f
1
O
S
O
8a2, 9a2, 10a2: R = Et
8a3, 9a3, 10a3: R1 = i-Bu
N
R1
HN
OCH3
NO2
8
7-95
Scheme 1. Reagents and conditions: (a) (i) NaNO , HCl, 10 ℃, (ii) SnCl , HCl, 30 ℃;(b) ethyl
2
2
pyruvate, EtOH, rt; (c) PPA, 80 ℃; (d) ArNH or aliphatic amines, Pd (dba) , DavePhos,
2
2
3
K PO (aq), toluene, 100℃; benzylboronic acid pinacol ester, Pd(dppf)Cl , K PO , CsF,
3
4
2
3
4
1,4-dioxane, H O, MW, 100 ℃; ArOH or ArSH, CuI, K CO , DME, MW, 130℃; (e) NaOH,
2
2
3
8
1
THF, EtOH and H O, r.t.; (f) R SO NH , HATU, DMAP, Et N, DCM, r.t.; (g) R X, Cs CO ,
2
2
2
3
2
3
DMF, r.t.
In order to explore the SARs, except for the nitro group on the 7-position of indole
ring, other substituents including F, Cl, CN, CF and CH were placed on either the
3
3
6
-position or 7-position of the indole ring. As depicted in Scheme 2, Hemetsberger
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
indole synthesis was applied to the synthesis of these derivatives. Upon the treatment
of ethyl 2-azidoacetate with substituted 2-bromobenzaldehyde (11a-11g) in the
presence of sodium ethoxide, 2-azido-acrylate derivatives (12a–12g) were produced
in 12-27% yields and later refluxed in xylene or 1,2-dichlorobenzene, forming the key
indole intermediates (13a–13g) in 30-93% yields. The coupling reaction of 13a–13g
with 3-methoxyaniline gave esters 14a–14g, which were then hydrolyzed into
indole-2-carboxylic acid derivatives (15a–15g). Subseqeuntly, they condensed with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
-methoxybenzenesulfonamide to access target molecules 80-86. The key
intermediate (18e) was produced in three steps in high yields, and its condensation
with various arylsulfonamides provided the N-methyl-7-chloro-substituted target
compounds 96-113. The 4-chloro-substituted target molecules 114 and 115 were
synthesized by using indole compound 13h as the starting material, and target
compound 79 was prepared starting from commercially available indole derivative
13i. Compound 96 was converted into its sodium salt (compound 118, Cpd118) in the
presence of NaOH in a quantitative yield.
OCH3
Br
O
Br
HN
Br
O
O
c
O
a
OEt
b
H
N3
2a-12g
N
OEt
N
OEt
R
R
R
R
H
H
1
14a-14g
11a-11g
13a-13g
OCH3
OCH3
11a-15a: R = 4-F
1b-15b: R = 4-Cl
11c-15c: R = 4-Me
11d-15d: R = 5-F
1
e
d
HN
HN
O
O
1
1
1e-15e: R = 5-Cl
1f-15f: R = 5-CN
O
S
O
N
H
OH
N
HN
0-86
OCH3
R
R
11g-15g: R = 5-CF3
H
15a-15g
8
ACS Paragon Plus Environment

Page 11 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
OCH3
OCH3
OCH3
Br
Cl
Br
Cl
HN
Cl
HN
Cl
HN
Cl
O
O
O
O
O
f
c
d
e
O
O
N
H
OEt
N
R8
OEt
N
OEt
N
OH
N
HN
S
1
3e
16e
17e
18e
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
96-113
OCH3
Br
Cl
Br
Cl
NH
Br
H
H
N
O
a
N3
b
N
O
c
H
OEt
OEt
OEt
O
O
Cl
Cl
1
1h
1
2h
13h
14h
OCH3
OCH3
NH
e
NH
d
H
N
H
O
N
O
O
S
O
OH
HN
OCH3
Cl
Cl
1
5h
114
OCH3
OCH3
OCH3
Br
Cl
Br
Cl
NH
NH
NH
H
N
O
N
O
N
O
N
O
N
O
f
c
e
d
O
S
OEt
OEt
OEt
OH
HN
OCH3
O
Cl
Cl
Cl
13h
16h
17h
18h
115
OCH3
OCH3
OCH3
Br
HN
HN
HN
O
O
O
c
d
O
e
O
N
OEt
N
H
N
H
OEt
OH
N
HN
S
OCH3
H
H
O
13i
14i
15i
OCH3
79
OCH3
HN
Cl
HN
g
O
HN
O
O
O
O
S
N
S
OCH3
N
N
OCH3
O
Cl
Na
9
6
118
Scheme 2. Reagents and conditions: (a) NaOEt, ethyl azidoacetate, EtOH; (b) xylene or
,2-dichlorobenzene, reflux; (c) 3-methoxyaniline, Pd (dba) , Xantphos, Na CO , toluene, H O,
1
2
3
2
3
2
8
reflux; (d) NaOH, THF, EtOH and H O, r.t.; (e) R SO NH , HATU, DMAP, Et N, DCM, r.t.; (f)
2
2
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
MeI, Cs CO , DMF, r.t.; (g) NaOH, H O, 80 ℃.
2
3
2
Compounds 116 and 117 bearing benzo[d]imidazole and azaindole scaffold
respectively, were also designed and synthesized as shown in Scheme 3. The
benzo[d]imidazole derivative 24 was built in a 5-step process starting from
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-nitrobenzene-1,2-diamine in 31% total yield. Then, starting from the key
intermediates 24 and 27 (commercially available), target molecules 116 and 117 were
constructed by undergoing palladium catalyzed coupling, base catalyzed hydrolysis
and condensation reaction.
NO2
NO2
NH2
NO2
NH2
NH2
Cl
Cl
Cl
N
O
N
N
O
d
N
N
O
N
c
a
b
N
H
N
H
OEt
OEt
OEt
NO2
19
20
21
22
23
OCH3
OCH3
OCH3
NH2
Cl
HN
Cl
HN
HN
e
O
f
N
N
O
g
N
N
O
h
N
O
N
O
O
N
OEt
OEt
OH
N
HN
S
OCH3
Cl
Cl
24
25
26
116
OCH3
OCH3
OCH3
Cl
N
HN
N
HN
N
HN
N
i
O
g
O
h
O
O
O
S
O
N
OEt
N
H
OEt
N
H
OH
N
HN
OCH3
H
H
29
27
28
117
Scheme 3. Reagents and conditions: (a) ethyl 2,2,2-trichloroacetimidate, TFA, ethyl ether/DCM;
b) Cs CO , EtOH, reflux; (c) MeI, Cs CO , DMF, r.t.; (d) 10% Pd/C, H , EtOH, 1atm, r.t.; (e)
(
2
3
2
3
2
NCS, DMF, r.t. (f) 3-bromoanisole, Pd (dba) , DavePhos, K PO (aq), toluene, 100 ℃; (g) NaOH,
2
3
3
4
THF, EtOH and H O, r.t.; (h) 4-methoxybenzenesulfonamide, HATU, DMAP, Et N, DCM, 40 ℃;
2
3
(i) 3-methoxyaniline, Pd (dba) , Xantphos, Na CO , toluene, H O, reflux.
2
3
2
3
2
ACS Paragon Plus Environment

Page 13 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Investigations on the SAR. Recombinant human FBPase activity was evaluated by
employing the coupling enzymes phosphoglucose isomerase and glucose-6-phosphate
+
dehydrogenase. The concomitant reduction of NADP to NADPH was monitored
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
spectrophotometrically.40 All target compounds (30-118) were tested for their
inhibitory activities against human liver FBPase. The corresponding results were
summarized in Tables 1-7. AMP and MB05032 were used as reference molecules.
Initially, a number of N-sulfonyl indole-2-carboxamide derivatives with different
8
R substituents were tentatively designed and tested. As shown in Table 1, the simple
methylsulfonamide 30 showed no inhibition against FBPase, while the
cyclopropylsulfonamide 31 had moderate inhibition (IC = 2.90 μM). In contrast, the
5
0
benzenesulfonamide 32 exhibited a much stronger potency with an IC value of 0.14
5
0
μM, which was 182-fold more potent than the hit compound A (IC = 25.6 μM). On
5
0
the basis of this result, several other representative aromatic sulfonamides (33-42)
were further evaluated. In fact, all of these aromatic sulfonamides including methoxyl,
fluoro, nitro-substituted benzene sulfonamides, thiophene sulfonamide, and
naphthalene sulfonamide exhibited high potency, with the IC values ranging from
5
0
0
.10 μM to 0.32 μM. Therefore, the above results evidently demonstrated that the
N-acyl arylsulfonamide was a critical structural feature of this series of inhibitors for
producing strong inhibition against FBPase. It was thus speculated that the N-acyl
sulfonamide moiety could form a hydrogen binding network with the key amino acids
of FBPase.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. The SAR of various N-acyl sulfonamide derivatives (30-42) against FBPasea
OCH3
A
NH
O
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R8
N
HN
S
H
O
NO2
Compd
30
R8
Me
IC50 (μM)
>50
Compd
37
R8
IC50 (μM)
0.16±0.01
0.10±0.01
0.21±0.03
0.27±0.08
0.32±0.01
0.28±0.02
4-fluorophenyl
3-nitrophenyl
4-nitrophenyl
3
3
1
2
cPr
2.90±0.80
0.14±0.01
0.15±0.06
0.19±0.03
0.24±0.09
0.14±0.00
38
Ph
39
33
3-methoxyphenyl
4-methoxyphenyl
2-fluorophenyl
3-fluorophenyl
40
4-(trifluoromethoxy)phenyl
thiophen-2-yl
3
4
41
35
42
naphthalen-2-yl
3
6
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
MB05032 was 0.044±0.012 μM.
As the aromatic sulfonamides were preferred and all of the tested compounds
(32-42) showed similar potency, we arbitrarily selected the un-substituted benzene
and 4-methoxybenzene as the B ring for subsequent SAR investigations on the
4
4-substituents (R ) of the indole scaffold (Table 2). The replacement of NH in
compound 34 with O and S atom produced compounds 43 and 44, which respectively
showed 7.3-fold and 3.7-fold decrease in activity. Furthermore, using the NCH (45,
3
IC > 50 μM) or the carbonyl group to substitute for NH (46, IC > 10 μM) resulted
50
50
in a loss of activity. Compounds with a two-atom linker were also designed. The
ACS Paragon Plus Environment

Page 15 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3-methoxybenzylamino-substituted compound 47 and benzamido-substituted
compound 48 presented potent inhibition as well. However, the inhibition reduced
about 5 times in comparison with compound 34. Putting together, these data
suggested that the NH group between the two aromatic rings was important for
inhibition. Presumably, this NH group served as a key H-bond donor in forming a
critical hydrogen bond with FBPase.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The alkylamino and cyclohexylamino groups were also incorporated into the
4-position of the indole ring, and the obtained compounds (49-52) displayed reduced
inhibition with IC values at a single digit micromolar level. When a morpholino
5
0
group or a chloro atom was installed onto the 4-position, the resultant compounds 53
IC = 35.6 μM) and 54 (IC = 24.3 μM) displayed an even lower potency. Hence,
(
5
0
50
these data demonstrated that an aromatic amino substituent on the 4-position of the
indole ring could strongly bolster the potency. It was conceived that the occurred
aromatic hydrophobic interaction and the hydrogen binding interaction have made
great contributions for the binding affinity. Therefore, at this stage, the scaffold of 4-
arylamino-N-arylsulfonyl-1H-indole-2-carboxamide was determined as the template
structure, which was subjected to further extensive SAR explorations.
Table 2.The SAR of 4-substituted indole derivatives (43-54) against FBPasea
R4
O
O
R10
N
HN
S
B
H
O
NO2
R4
R10
IC (μM)
50
Compd
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
O
O
O
3
4
4
4
3
4
OMe
OMe
OMe
0.19±0.01
1.40±0.22
0.70±0.02
NH
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
S
4
4
5
6
OMe
OMe
OMe
>50
>10
N
O
O
47
0.97±0.03
HN
O
4
4
5
8
9
0
OMe
OMe
OMe
OMe
OMe
H
0.87±0.02
1.40±0.80
1.70±0.00
1.80±0.10
2.00±0.25
35.6±5.10
24.3±3.20
HN
H
N
H
N
O
H
O
N
51
NH
5
5
5
2
3
4
O
N
Cl
H
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
MB05032 was 0.044±0.012 μM.
The effects of variations on the A ring were evaluated and outlined in Table 3. The
mono-substituted A ring with various substituents on different positions was first
investigated. Although all of these compounds (32, 34, 55-69) were active against
FBPase with IC50 ranging from 0.12 μM to 3.10 μM, the substituted pattern of
substituents had impacts on the inhibition. Introducing a methoxyl group on the
ACS Paragon Plus Environment

Page 17 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2-position would markedly deteriorate the activity, as compound 55 (IC = 3.10 μM)
5
0
was the least potent derivative among compounds 32, 55 and 56, and it was 22-fold
less potent than compound 32 with a methoxyl group on the 3-position (IC = 0.14
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
μM). Furthermore, placing a moiety on the 3-position was usually more favorable
than that on the 4-position (32 vs 56, 62 vs 63, 68 vs 69). Also, a variety of moieties
were tolerated on the 3-position and most of them strongly inhibit the FBPase activity
with IC values ranging from 0.12 μM to 1.01 μM. Particularly, compounds with
5
0
3
-substituents such as OMe, OEt, OCF H, CH , NHCOCH , and NO displayed even
2 3 3 2
stronger inhibition in the series, and their IC values are of less than 0.16 μM.
5
0
Among the di-substituted compounds (70-74), 3,4- and 3,5-disubstituted
compounds 70, 73, and 74 were more active than 2,3- and 2,4-disubstituted
compounds 71 and 72. Hence, it was further supported that substitution on the
2-position of A ring was unfavorable for the activity. As for the 3,4,5-trisubstituted
derivative, compound 75 (IC = 0.20 μM) exhibited high potency as well.
5
0
Overall, the above-mentioned results suggested that the substitution on the
ortho-position of A ring was not tolerated. Presumably, the 2-substituent might bring
about a steric hindrance or interfere with the hydrogen bond formation of the adjacent
NH group. In contrast, the meta- and para-position could accommodate a variety of
groups, either electron-donating groups or electron-withdrawing moieties, and that
could provide a chance for further improving the structural diversity and drug-like
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
properties. Given consideration that grafting a substituent on the 3-position of A ring
significantly enhanced the inhibition and taking synthetic simplicity and drug-like
properties into account, we tentatively selected compound 34 as the lead for the next
round of optimization.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 3.The SAR of substitution on the A ring of 7-nitroindole derivatives (55-75) against
FBPasea
R9
A
NH
O
O
R10
N
HN
S
B
H
O
NO2
R9
R10
R9
R10
Compd
IC50 (μM)
Compd
65
IC50 (μM)
0.91±0.21
0.14±0.03
0.35±0.02
0.31±0.01
1.09±0.22
0.14±0.01
1.46±0.35
1.28±0.11
0.12±0.01
0.40±0.02
0.20±0.01
3
3
2
4
3-OMe
3-OMe
2-OMe
4-OMe
3-OEt
H
0.14±0.02
0.19±0.01
3.10±0.31
1.00±0.20
0.12±0.02
0.13±0.02
0.15±0.11
0.43±0.01
0.16±0.02
1.01±0.11
1.51±0.33
0.66±0.01
4-CF
3
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
H
66
3-NO
4-NO
2
2
55
67
5
5
5
5
6
7
8
9
H
68
3-morpholino
4-morpholino
3,5-dimethoxy
2,3-dimethoxy
2,4-dimethoxy
4-Cl-3-OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
69
3-OCF
2
H
70
3-Me
3-Et
71
60
72
6
6
6
6
1
2
3
4
3-acetamido
73
3-F
4-F
74
4-F-3-OMe
75
3,4,5-trimethoxy
3-CF
3
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
ACS Paragon Plus Environment

Page 19 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
MB05032 was 0.044±0.012 μM.
Modifications on the 5-, 6- and 7-positions of the indole ring were subsequently
conducted by taking compound 34 as the template structure, and the results were
listed in Table 4. Introducing an F atom on the 5-position (76, IC = 0.26 μM)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
produced a comparable activity to compound 34, while incorporating a Cl or Me
group resulted in a 5-fold decrease in activity, suggesting that a bulky substituent was
not allowed on the 5-position.
Removing the 7-nitro group led to a drastic decline in inhibition as compound 79
was about 20-fold less active than compound 34, indicating that the 7-substituent
plays a key role in the binding affinity. Therefore, we further prepared other
7
-substituted derivatives. Compounds (83, 85, 86) with an F, CN or CF group could
3
potently inhibit FBPase, and their IC values are of 0.57 μM, 0.36 μM and 0.51 μM,
5
0
respectively. It was noteworthy that compound 84 (IC = 0.16 μM) with a Cl atom
5
0
offered a potent and comparable activity to compound 34. It is well-known that the
NO group is not desirable in the drug molecule in terms of drug-like properties, and
2
thus we explored the 7-Cl substituted series of compounds in the later optimization,
leading to a candidate molecule Cpd118.
The incorporation of an F, Cl or CH group on the 6-position of the indole ring
3
afforded compounds 80-82, which displayed a reduced inhibition (IC , 1.1 μM - 3.6
5
0
μM), suggesting that the 6-substituent on the indole ring was not boosting the binding
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
affinity. Therefore, we retained a nitro group or a chloro atom on the 7-position of the
indole ring in the following SAR investigation.
Table 4. The SAR of substitution on the 5-, 6- and 7-positions of indole derivatives (76-86)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
against FBPasea
OCH3
A
NH
R5
R6
O
O
S
N
HN
B
OCH3
H
R7
O
Compd R5 R6
R7
IC (μM)
Compd R5 R6
R7
H
H
F
IC (μM)
5
0
50
3
4
H
F
H
H
H
H
H
F
NO2 0.19±0.02
NO2 0.26±0.01
NO2 0.95±0.01
NO2 0.95±0.02
81
82
83
84
85
86
H
H
H
H
H
H
Cl
Me
H
1.1±0.01
3.6±0.62
0.57±0.02
76
7
7
Cl
Me
H
78
H
Cl 0.16±0.02
CN 0.36±0.01
CF3 0.51±0.02
7
8
9
0
H
H
3.7±0.42
1.2±0.14
H
H
H
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
MB05032 was 0.044±0.012 μM.
The effects of alkylation on the indole nitrogen atom of the lead compound 34 were
also explored and illustrated in Table 5. The methylation on the nitrogen atom
produced a remarkable boost on inhibition, compound 87 presented double-digit
nanomolar potency with an IC value of 0.04 μM, which was comparable to the
5
0
clinical candidate MB05032. In contrast, introducing an ethyl or isobutyl group did
not increase the potency, the IC values of compounds 88 and 89 were 0.21 μM and
5
0
ACS Paragon Plus Environment

Page 21 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
0.40 μM, respectively. Clearly, the methyl group was more favorable than the larger
alkyl group, suggesting that a small pocket existed around the indole nitrogen. Thus,
the methylation was performed on a number of the previously obtained potent
derivatives with IC values of 0.10 μM - 0.20 μM. In fact, the N-methylation did not
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
always give rise to an enhancement in potency. Some of the compounds (90, 92, 93
and 95) produced an increased activity, and their IC values (0.059 μM – 0.09 μM)
5
0
-8
reached the 10 M level. In contrast, compound 91 (IC = 0.23 μM) showed
5
0
comparable activity to its unsubstituted counterpart; and compound 94 (IC = 1.30
5
0
μM) with a trimethoxyl group displayed a 6-fold decrease in potency compared with
the corresponding unsubstituted derivative 75. These facts implicated that the
N-methylation not only brought about a beneficial hydrophobic interaction, but also
might have an impact on the binding of substituted A ring.
Table 5. The SAR of N-alkylation of 7-nitroindole derivatives (87-95) against FBPasea
R9
A
NH
O
O
S
N
R1
HN
B
OCH3
O
NO2
R1
R9
R1
R9
Compd
34
IC50 (μM)
Compd
91
IC50 (μM)
0.23±0.01
H
3-OMe 0.19±0.01
Me
Me
Me
3-EtO
8
8
8
9
7
8
9
0
Me 3-OMe 0.04±0.01
Et 3-OMe 0.21±0.02
iBu 3-OMe 0.40±0.11
92
3-acetamido
3,5-dimethoxy
0.059±0.002
0.063±0.001
1.30±0.22
93
94
Me 3,4,5-trimethoxy
Me
3-Me
0.09±0.01
95
Me
4-Cl-3-OMe
0.063±0.003
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
MB05032 was 0.044±0.012 μM.
Since the 7-chloro-substituted indole derivatives were supposed to be more
drug-like than the 7-nitro substituted derivatives, we further broadened the SARs
based on this template. As shown in Table 6, the methylation of compound 84
afforded compound 96, which exhibited remarkable inhibition with an IC value of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
0
.052 μM, and this potency was comparable to that of compound 87 (the 7-nitro
substituted analog) and MB05032. The other four compounds (103, 105-107) with a
methyl, trifluoromethyl or cyano substituent on the meta- or para-position of benzene
ring were also identified with a pronounced potency (IC , 0.056 μM - 0.085 μM). Of
5
0
note, compound 104 (IC50 = 0.77 μM) with a trifluoromethyl group on the
ortho-position and compound 108 (IC = 0.94 μM) with a bulky iso-butyl group on
5
0
the meta-position of benzene ring showed moderate inhibition. In addition, the
benzo-heterocyclic derivatives (111-113) were also examined and none of them
showed the desired high potency. To summarize, these results suggested that the
binding pocket occupied by the arylsulfonyl moiety was a key binding site and an
appropriate modification on this fragment would provide an opportunity for the
improvement in potency.
Table 6. The SAR of N-acyl arylsulfonamide containing 7-chloroindole derivatives (96-113)
against FBPasea
ACS Paragon Plus Environment

Page 23 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
OCH3
OCH3
A
A
NH
Cl
NH
O
O
O
O
S
N
HN
S
R8
N
HN
R8
H
O
O
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cl
8
4
96-113
Compd
R8
IC (μM)
Compd
105
R8
IC (μM)
5
0
50
O
CF3
O
O
8
9
4
6
0.16±0.02
0.085±0.011
0.07±0.02
0.056±0.002
0.94±0.02
0.22±0.08
0.13±0.03
0.59±0.26
O
CF3
0.052±0.006
0.18±0.03
0.15±0.01
0.14±0.02
0.23±0.01
0.47±0.06
106
107
108
109
110
111
CN
97
O
9
9
8
9
F
N
F
S
1
00
F
101
N
Br
1
1
02
03
0.14±0.02
112
113
0.54±0.08
0.67±0.11
N
0.071±0.004
N
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
F3C
O
1
04
0.77±0.03
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
MB05032 was 0.044±0.012 μM.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As shown in Table 7, we also exchanged the position of the chloro atom and
3-methoxyphenylamino group of compounds 84 and 96 resulted in 4-chloro
substituted compounds 114 and 115. Compound 114 (IC = 0.10 μM) had similar
5
0
potency with compound 84, while the inhibition of the N-methyl derivative 115
IC > 10 μM) was declined markedly, indicating that the methyl group on this
(
5
0
position could severely interfered with the binding due to its steric hindrance,
presumably. In comparison with compound 96, as the close analogs of indole
scaffold, the benzo[d]imidazole analog 116 (IC = 0.027 μM) offered a comparable
5
0
inhibition, and the aza-indole analog 117 showed a 15-fold reduction in potency. The
replacement of the indole scaffold would be a choice for obtaining novel FBPase
inhibitors with diversified structures.
Table 7. The SAR of various heterocyclic rings containing derivatives (114-118) against
FBPasea
Compd
structure
IC (μM)
50
OCH3
1
14
0.10±0.01
NH
H
N
O
O
O
HN
S
OCH3
Cl
ACS Paragon Plus Environment

Page 25 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
OCH3
NH
115
>10
N
O
O
O
HN
S
OCH3
Cl
OCH3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH
Cl
1
16
0.027±0.003
N
N
O
O
O
HN
S
OCH3
OCH3
NH
N
117
0.71±0.02
O
O
S
O
N
H
HN
OCH3
OCH3
0
.029±0.006
NH
1
18
O
O
O
N
N
S
OCH3
Cl
Na
aAMP and MB05032 were used as reference molecules. IC50 for AMP was 3.3±0.1 μM. IC50 for
MB05032 was 0.044±0.012 μM.
The Analysis of Cocrystal Structure of FBPase Complexed with Cpd118. The
X-ray cocrystal structure of Cpd118 bound to FBPase was solved to a resolution of
2
.40 Å (PDB: 6LW2). As shown in Figure 3, the tetramer of the protein contained
four small molecules, and each molecule occupied two sites, the adenine pocket of
AMP binding site and the interface region between two monomers (Figure 3, A and
B). Obviously, Cpd118 has a very different binding mode from AMP, which fully
occupied both the adenine sub-pocket and the phosphate binding site. As shown in
Figure 3C and 3D, Cpd118 occupied the adenine binding pocket with its
4
-methoxybenzenesulfonamide fragment interacting with amino acids Glu20, Ala24,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and Leu30 through hydrophobic interactions, the N-acyl sulfonamide linker, which
was tentatively used as the bioisostere of the phosphate, was situated above of the
phosphate pocket. Noticeably, a water molecule was observed in this phosphate
binding site instead, bridging rich hydrogen binding interactions between the O atom
of the sulfonyl group and the characteristic amino acids Thr27, Glu29, Lys112, and
Tyr113. These binding interactions would certainly boost the binding affinity.
Importantly, the N-acyl sulfonamide linker itself actually formed an extensive
hydrogen bond network with the key amino acids Leu30 and Thr31, which were
supposed to contribute to the binding affinity to a great extent. Furthermore, these
H-bond interactions facilitated this linker to protrude into the narrow channel formed
by Gly26, Thr27, and Gly28 on the front and Leu30 and Thr31 on the back, thus
orientating the indole part into the subunit interface region (Figure 3, C). Strikingly,
at this interface area, the extensive π-π stacking interactions were observed: one is
between the two indole rings, and the other is between the two 3-methoxybenzene
rings from monomers A (B) and D (C), respectively. These strong aromatic
hydrophobic interactions were another critical feature of this series of inhibitors, and
played remarkable roles for stabilization of the complex. Moreover, the indole ring
interacted with the side chain of Arg22, and the 3-methoxyphenyl fragment interacted
with Arg25 through hydrophobic interactions as well. Additionally, one more critical
binding feature was also observed. The NH group between the two aromatic moieties
formed a hydrogen bond with the C=O group of Gly26. This was consistent with the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
results of SARs as a replacement of the NH group with NCH would lead to a marked
3
ACS Paragon Plus Environment

Page 27 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
decrease in potency. It was also observed that both N-methyl group and 7-chloro atom
on the indole ring situated in a hydrophobic area lined with the side chains of Met18
and Arg22, and the SARs demonstrated that these small hydrophobic groups had
pronounced effects on the binding affinity.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. The cocrystal structure of FBPase in complex with Cpd118 (PDB 6LW2) (A)
Overview of FBPase comprised of four monomers and a close view of the ligands; (B) The
binding pose of Cpd118 which occupied the subunits interface and a part of AMP binding
site (the AMP from 1FTA was shown in red);36 (C) The binding pose of Cpd118 and the
tunnel formed by Gly26, Thr27 and Gly 28; (D) Detailed interactions of Cpd118 with key
amino acides, green dashes and blue dashes for H-bonds and purple dashes for hydrophobic
interactions, the distance for H-bonds and the centroid of aromatic rings were labelled. The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 91
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
image was created with PyMol software.41 The 2F -F electron density maps of (A) and (B)
o
c
were contoured at 1σ.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Collectively, Cpd118 took a distinct binding mode, which simultaneously
exploited both the AMP domain and the interface region via three characteristic
fragments including the N-acyl sulfonamide, 4-methoxybenzene ring, and
4
-arylaminoindole moiety. These structural features created a very different binding
mode from that of the well-known AMP mimics such as MB05032, which fully made
use of the AMP allosteric binding site. Interestingly, we noted that the benzoxazole
benzenesulfonamide derivatives (Figure 1) and the sulfonyl ureido derivatives
(Figure 1) were reported to have a similar binding mode to Cpd118.20,21,30,31 They all
utilized the adenine pocket of AMP and the hydrogen binding interactions with Leu30
and Thr31 as well. However, at the interface binding region, these two types of
molecules did not build up extensive π-π stacking interactions as Cpd118. In
summary, this series of N-acyl arylsulfonamide inhibitors with a unique binding mode
provided an opportunity for achieving highly potent and drug-like allosteric dual
binding site inhibitors against FBPase.
The Enzymatic Activity and Selectivity of Cpd118. The pyruvate is a substrate of
the gluconeogenesis process. The oral administration of pyruvate to fasting animals
can result in an enhancement of blood glucose level through the gluconeogenesis
process. Therefore, the oral pyruvate tolerance test (OPTT) in normal fasting ICR
ACS Paragon Plus Environment

Page 29 of 91
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mice or alloxan-induced diabetic mice was used to quickly screen the in vivo efficacy
of the obtained highly potent FBPase inhibitors, which had IC values at the double
5
0
digit nanomolar level. As a result, Cpd118 (the sodium salt of compound 96) was
found to be the most potent compound in the OPTT test by measuring the blood
glucose level. Hence, Cpd118 was selected for further evaluation in terms of its
enzymatic activity, selectivity, and in vivo glucose lowering efficacy in type 2 diabetic
animal models.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As shown in Table 8, Cpd118 exhibited strong inhibition against FBPase
(0.029±0.006 μM), whereas no significant inhibition or activation was observed
towards six other key AMP-binding enzymes with their IC or EC values of more
5
0
50
than 100 μM or 10 μM, respectively. Cpd118 displayed an excellent selectivity
against all six tested enzymes.
Table 8. FBPase inhibitory potency and selectivity of Cpd118
a
a
Enzymes
IC (μM)
EC (μM)
5
0
50
Human liver FBPase
0.029±0.006
> 100
> 100
> 100
—
—
—
AMP deaminase
Adenylate kinase (AK)
Adenosine kinase (ADK)
Glycogen phosphorylase
Phosphofructokinase (PFK)
AMP-activated protein kinase
—
—
> 100
> 100
> 10
—
—
ACS Paragon Plus Environment