Design, Synthesis, and Preclinical Efficacy of Novel Nonretinoid Antagonists of Retinol-Binding Protein 4 in the Mouse Model of Hepatic Steatosis

Article abstract of DOI:10.1021/acs.jmedchem.9b00352

Retinol-binding protein 4 (RBP4) serves as a transporter for all-trans-retinol (1) in the blood, and it has been proposed to act as an adipokine. Elevated plasma levels of the protein have been linked to diabetes, obesity, cardiovascular diseases, and nonalcoholic fatty liver disease (NAFLD). Recently, adipocyte-specific overexpression of RBP4 was reported to cause hepatic steatosis in mice. We previously identified an orally bioavailable RBP4 antagonist that significantly lowered RBP4 serum levels in Abca4^-/- knockout mice with concomitant normalization of complement system protein expression and reduction of bisretinoid formation within the retinal pigment epithelium. We describe herein the discovery of novel RBP4 antagonists 48 and 59, which reduce serum RBP4 levels by >80% in mice upon acute oral dosing. Furthermore, 59 demonstrated efficacy in the transgenic adi-hRBP4 murine model of hepatic steatosis, suggesting that RBP4 antagonists may also have therapeutic utility for the treatment of NAFLD.

Full text of DOI:10.1021/acs.jmedchem.9b00352

Subscriber access provided by University of Groningen
Article
Design, Synthesis, and Pre-Clinical Efficacy of
Novel Non-Retinoid Antagonists of Retinol Binding
Protein 4 in the Mouse Model of Hepatic Steatosis
Christopher L. Cioffi, Boglarka Racz, András Váradi, Emily Freeman, Michael P
Conlon, Ping Chen, Lei Zhu, Douglas B. Kitchen, Keith Barnes, William Martin,
Paul Pearson, Graham Johnson, William S. Blaner, and Konstantin Petrukhin
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00352 • Publication Date (Web): 11 May 2019
Downloaded from http://pubs.acs.org on May 12, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Design, Synthesis, and Pre-Clinical Efficacy of Novel Non-Retinoid
Antagonists of Retinol Binding Protein 4 in the Mouse Model of
Hepatic Steatosis
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
€
§
§
Christopher L. Cioffi,* Boglarka Racz, Andras Varadi, Emily E. Freeman,† Michael P.
&
&
‡
Conlon,† Ping Chen,† Lei Zhu, Douglas B. Kitchen, Keith D. Barnes,† William H. Martin,
¶
£
¥
§
Paul G. Pearson, Graham Johnson, William S. Blaner and Konstantin Petrukhin*
€Albany College of Pharmacy and Health Sciences, Departments of Basic and Clinical
Sciences and Pharmaceutical Sciences, 106 New Scotland Ave, Albany, NY 12208
§Department of Ophthalmology, Columbia University Medical Center, New York, NY
1
0032
†AMRI, Department of Medicinal Chemistry, East Campus, 3 University Place, Rensselaer,
NY 12144
&AMRI, Computer Assisted Drug Discovery, East Campus, 3 University Place, Rensselaer,
NY 12144
‡WHM Consulting LLC, 111 Sterling City Road, Lyme, CT 06371
¶
Pearson Pharma Partners, 31194 La Baya Drive, Westlake Village, CA 91361
£NuPharmAdvise LLC, 3 Lakeside Drive, Sanbornton, NH 03269
¥Department of Medicine, Columbia University Medical Center, New York, NY 10032
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 112
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract: Retinol-binding protein 4 (RBP4) serves as a transporter for all-trans retinol (1)
in the blood and it has been proposed to act as an adipokine. Elevated plasma levels of
the protein have been linked to diabetes, obesity, cardiovascular diseases, and non-
alcoholic fatty liver disease (NAFLD). Recently, adipocyte-specific overexpression of RBP4
was reported to cause hepatic steatosis in mice. We previously identified an orally
bioavailable RBP4 antagonist that significantly lowered RBP4 serum levels in Abca4-/-
knockout mice with concomitant normalization of complement system protein
expression and reduction of bisretinoid formation within the retinal pigment epithelium
(RPE). We describe herein the discovery of novel RBP4 antagonists 48 and 59, which
reduce serum RBP4 levels by >80% in mice upon acute oral dosing. Furthermore, 59
demonstrated efficacy in the transgenic adi-hRBP4 murine model of hepatic steatosis,
suggesting that RBP4 antagonists may also have therapeutic utility for the treatment of
NAFLD.
0
1
2
3
4
5
6
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
Retinol-binding protein 4 (RBP4) is a serum protein belonging to the lipocalin family that
transports the essential vitamin all-trans-retinol (vitamin A, 1) (Figure 1) from the hepatic
1
stores to peripheral tissues. RBP4-mediated transport of 1 from the liver to target tissues
requires formation of a tertiary complex between holo-RBP4 (RBP4 bound to 1) and
2
transthyretin (TTR) as circulating apo-RBP4 (RBP4 not bound to 1) associates poorly with
TTR and due to its low molecular weight (21 kDa), is rapidly cleared via glomerular
3
filtration. Inhibition of holo-RBP4 formation by displacing 1 from the ligand-binding
2
ACS Paragon Plus Environment

Page 3 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pocket of RBP4 by selective antagonists prevents RBP4-TTR complexation from occurring,
which in turn leads to a reduction in circulating levels of RBP4 and 1. Cellular uptake of 1
is facilitated by STRA6 (stimulated by retinoic acid gene 6 protein), which functions as a
high affinity RBP4 membrane receptor and a retinol channel in many vitamin A-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
dependent tissues and blood-organ barriers. Until recently, transport of 1 was the only
function attributed to RBP4. Consistent with its major role as a carrier of hepatic 1, 60-
5
80% of circulating RBP4 derives from the liver where it is secreted bound to 1 with extra-
hepatic secretion, mainly from adipocytes, accounting for approximately 20-40% of
circulating levels of the protein.6-8 Recently, emerging epidemiologic evidence suggested
that modestly enhanced circulating levels of RBP4 might have important metabolic
effects. Numerous, but not all, epidemiological studies established an association of
elevated serum RBP4 levels with obesity,8,
8
9
insulin resistance,8-11 cardiovascular
disease,12, 13 pro-atherogenic conditions, hepatic steatosis, and type 2 diabetes.
9
8, 11
It
has been theorized that RBP4 may act as an adipokine (a cytokine produced by
adipocytes) involved in the mediation of insulin sensitivity and the regulation of energy
2
metabolism. Consistent with this hypothesis, it was reported that RBP4 secreted from
cultured adipocytes may activate cytokine production from co-cultured macrophages
through TLR4 and JNK pathway activation that does not involve the RBP4 receptor
STRA6.13, 14 It was suggested that retinol binding may not be required for RBP4 to act as a
pro-inflammatory metabolic effector, given that apo-RBP4 was shown to induce an
inflammatory response in endothelial cells by stimulating the expression of pro-
inflammatory signaling molecules in a retinol- and STRA6-independent fashion.15
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 112
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
Conversely, it has been reported that RBP4 signaling depends on retinol binding and
involves the STRA6 receptor.16, 17 Despite the progress made in the characterization of
RBP4 as a putative adipokine, it is not fully understood how modestly elevated serum
RBP4 levels in humans may mechanistically contribute to the pathogenesis of a variety of
metabolic disorders. Three mouse models mimicking the increase in circulating levels of
serum RBP4 were used to probe the role of RBP4 in mediating adverse metabolic
consequences. Transgenic mice ectopically expressing human RBP4 in muscle from a
muscle-specific promoter have approximately 3-fold higher circulating levels of serum
0
1
2
3
4
5
6
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
RBP4.18 These mice were reported to develop high-fat diet (HFD)-induced insulin
9
resistance, although two subsequent studies could not replicate this finding.19, 20 Adeno-
associated virus-driven overexpression of murine RBP4 in the mouse liver yielded a 2- to
21
3
-fold increase in levels of serum RBP4. However, no effect on glucose and energy
homeostasis was detected in response to this significant increase in liver-secreted
circulating RBP4.21 In a remarkable contrast to the two previous models, low-level
transgenic expression of human RBP4 in the mouse adipose tissue led to a strong
metabolic phenotype, which comprised hepatic steatosis, dyslipidemia, and reduced
2
2
glucose tolerance. Given that hepatic steatosis observed in this model resulted from
2
2
increased hepatic uptake of adipose-derived circulating non-esterified fatty acids, it may
be suggested that adipose-secreted RBP4 facilitates trafficking of the adipose-derived
fatty acids to the liver through a yet to be defined mechanism. Interestingly, earlier crystal
structures of heterologously expressed RBP4 display fortuitous fatty acid ligands from the
expression host in the ligand binding pocket of RBP4.23, 24 Furthermore, recent x-ray
4
ACS Paragon Plus Environment

Page 5 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
crystallographic and mass spectrometry findings reported by Monaco and coworkers
confirmed that in addition to binding retinoids, RBP4 is also capable of binding fatty
acids,25 which may potentially relate to the RBP4 role as a facilitator of fatty acid
trafficking. We previously described the discovery new classes of non-retinoid RBP4
antagonists26, 27 that we studied in animal models of retinal disease. In light of the
aforementioned studies, it seems reasonable to suggest that synthetic RBP4 ligands may
also serve as important pharmacological tools in assessing the role of RBP4 in the
pathogenesis of metabolic disorders such as non-alcoholic fatty liver disease (NAFLD) and
non-alcoholic steatohepatitis (NASH). Here we report the characterization of a new class
of non-retinoid RBP4 antagonists and their assessment in the mouse transgenic model of
hepatic steatosis.
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
28
Protein Data Bank (PDB) high-resolution x-ray crystal structures providing binding
poses of various ligands bound to RBP4 were studied during our structure-based drug
design efforts to discover novel non-retinoid RBP4 antagonists. These structural data
include RBP4 co-crystallized with 1 (PDB 1rbp), 1 bound to the RBP4-TTR tertiary complex
(PDB 3bsz), RBP4 co-crystallized with the retinoid RBP4 antagonist fenretinide (2) (PDB
1
fel), and RBP4 co-crystallized with the non-retinoid RBP4 antagonist A1120 (3) (PDB
3fmz).26, 27 Initial rational design efforts derived from the aforementioned structural data
provided optimized bicyclic antagonists 4 and BPN-14136 (5). Importantly, 5 presented
exceptional in vitro RBP4 binding and functional antagonist activity, desirable ADME
characteristics, and a favorable pharmacokinetic (PK) profile. In addition, 5 dramatically
reduced circulating RBP4 levels in vivo with good pharmacokinetic/pharmacodynamic
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 112
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
correlation (PK/PD) in rodents. Furthermore, 5 significantly reduced lipofuscin bisretinoid
accumulation and normalized complement system protein expression in the retinal
-/-
pigment epithelium (RPE) of Abca4 mice, a model that recapitulates the phenotype 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
Stargardt disease.28
Figure 1. All-trans-retinol or vitamin A (1), the retinoid RBP4 antagonist fenretinide (2), the non-
retinoid piperidine-based RBP4 antagonist A1120 (3), and the rationally designed bicyclic
analogues 4 and BPN-14136 (5).
Buoyed by the in vitro and in vivo profiles of 5, we sought to further build upon our
findings and expand our compound portfolio by developing additional structurally diverse
series of RBP4 antagonists. Our objective was to identify novel chemical matter devoid
of a carboxylic acid. Carboxylic acids are present on numerous drugs and our earlier RBP4
antagonists utilized this functional group effectively. The carboxylic acid can confer
numerous benefits such as reducing CNS exposure, however some carboxylic acid bearing
6
ACS Paragon Plus Environment

Page 7 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds have also been reported to present idiosyncratic drug toxicity liabilities that
2
9
have been associated with acyl glucuronidation formation. Thus, we sought to further
expand our chemical matter portfolio by developing diverse series that could avoid
potential acyl glucuronidation liabilities should they arise. Drawing from the reported N-
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
30
benzyl imidazole RBP4 antagonist 6 (hRBP4 SPA IC50 = 0.44 ± 0.18 M) (Figure 2) and
guided by our previously reported 3fmz-derived computational docking model,26, 27 we
implemented a structure-activity relationship (SAR) strategy involving the design and
synthesis of novel compounds presenting 2- and 3-carboxamido fused [5,6]-bicyclic
heteroaromatic “bottom group” appendages. Our goal was to reduce the number of
rotatable bonds of 6 by fusing the pendent benzyl group phenyl ring directly to the
imidazole ring via a [5,6]-bicyclic heteroaromatic framework. We hypothesized that
motifs that remove these rotatable bonds but still project an aromatic ring toward the
opening of the binding pocket may provide novel chemical matter with improved potency
and physicochemical properties relative to 6.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 112
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. N-benzyl imidazole RBP4 antagonist 6 and the SAR strategy used to discover novel RBP4
antagonists containing 2- and 3-carboxamido fused [5,6]-bicyclic heteroaromatic appendages.
Figure 3 highlights how we initially executed our medicinal chemistry work plan. Novel
analogues were synthesized with the 4-(2-(trifluoromethyl)phenyl)piperidine core
scaffold of 3 and in some cases with the bicyclic core system of 4 and 5. We began our
campaign by exploring 3-carboxamides of various fused bicyclic heteroaromatic systems
appended to both cores followed by subsequent syntheses of analogues bearing 2-
carboxamido fused bicyclic heteroaromatic systems. Select piperidine core analogues
bearing a 3-carboxamido [1,2,4]triazolo[4,3-a]pyridine-6-carbonitrile bottom group motif
with alternatively fluorinated 2-trifluoromethylphenyl head groups were also later
prepared.
8
ACS Paragon Plus Environment

Page 9 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. The structure-based drug design strategy used to discover novel RBP4 antagonists
containing 3- and 2-carboxamido fused [5,6]-bicyclic heteroaromatic appendages starting from
lead 6. (A) The piperidine (A) and bicyclic octahydrocyclopenta[c]pyrrole (B) scaffold core systems
used in the SAR campaign. (B) The initial 3-carboxyamido and subsequent 2-carboxyamido fused
[
5,6]-bicyclic heteroaromatic bottom group motifs explored.
.
CHEMISTRY
The syntheses of compounds possessing 1H-indazole-3-carboxamides (11–18) are
31
presented in Scheme 1. Tertiary alcohol 8 is manufactured via addition of lithiated 7 to
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 112
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
N-benzyl-4-piperidinone. Elimination of 8 upon treatment with thionyl chloride yields
tetrahydropyridine 9, which was reduced and N-benzyl de-protected in one-pot via a
Pd/C-catalyzed ammonium formate reduction. Treatment of the resultant piperidine
freebase with HCl provided hydrochloride salt 10, which underwent facile amide
formation with various 1H-indazole-3-carboxylic acids to generate amide analogues 11–
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
8 using HBTU as the peptide coupling agent.
Scheme 1a
aReagents and conditions: (a) (i) n-BuLi, THF, -78 °C, 40 min; (ii) 1-benzylpiperidin-4-one, THF, -78
°
2
C; (b) SOCl , 0 °C, 2 h; (c) (i) HCO
2 4
NH , 10% Pd/C, CH
3
OH, reflux, 2 h; (ii) 4.0 M HCl solution in 1,4-
3
dioxane, CH CN, rt, 2 h; (d) substituted indazole-3-carboxylic acid, HBTU, Et
3
N, DMF, rt, 16 h.
10
ACS Paragon Plus Environment

Page 11 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The syntheses of N1-substituted-1H-indazole-3-carboxamides 19–21 are highlighted
in Scheme 2. Alkylation of 1H-indazole 13 with either iodoethane or 2-iodopropane gave
1
9 and 20, respectively. Formation of desired oxetane 21 was achieved via alkylation
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
between 13 and iodooxetane.
Scheme 2a
aReagents and conditions: (a) iodoethane or 2-iodopropane, K
CO , DMF, rt, 4h; (b) 3-
2
3
iodooxetane, K
2
CO , DMF, 60 °C, 24 h.
3
The syntheses of benzo[d]isoxazole-3-carboxamide 22 and benzo[c]isoxazole-3-
carboxamide 23 are depicted in Scheme 3. HBTU-facilitated peptide coupling of
benzo[d]isoxazole-3-carboxylic acid or benzo[c]isoxazole-3-carboxylic acid with
piperidine hydrochloride 10 afforded 22 and 23, respectively.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 112
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
Scheme 3a
0
1
2
3
4
5
6
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
aReagents and conditions: (a) benzo[d]isoxazole-3-carboxylic acid, HBTU, Et
benzo[c]isoxazole-3-carboxylic acid, HBTU, Et N, DMF, rt, 16 h.
3
N, DMF, rt, 16 h; (b)
3
We previously disclosed the synthesis of bicyclic intermediate 31, which is highlighted
32
in Scheme 4. Reduction and subsequent N-Boc protection of dione 24 gave unsaturated
isoindole 25, which underwent oxidative cleavage to provide di-acid 26. Dieckman
condensation followed by de-carboxylation gave ketone intermediate 27, which was
converted to racemic vinyl triflate (±)-28. Compound (±)-28 underwent Pd-catalyzed
Suzuki cross-coupling reaction to give styrene (±)-29, which underwent Pd/C-catalyzed
hydrogenation to provide endo-isomer 30 exclusively. N-Boc de-protection with 2.0 M HCl
in Et O afforded hydrochloride salt 31.
2
12
ACS Paragon Plus Environment

Page 13 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 4a
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
aReagents and conditions: (a) (i) LiAlH
rt, 16 h; (b) (i) NaIO , RuO2•H O, CH CN, CCl
LiHMDS (1.0 M solution in THF), THF, -78 °C, 30 min; (ii), PhN(SO
2-(trifluoromethyl)phenyl)boronic acid, Pd(PPh , 2.0 M Na CO , DME, 80 °C, 6 h; (f) H
0% Pd/C, CH OH, rt, 16 h; (g) 2.0 M HCl in Et O, CH Cl , 0 °C to rt, 24 h.
(1.0 M solution in THF), THF, 70 °C, 16 h; (ii) Boc
, H O, rt, 24 h; (c) Ac O, NaOAc, 120 °C, 3 h; (d) (i)
CF , THF, -78 °C to rt, 3 h; (e)
(40 psi),
O, CH
Cl ,
4
2
2
2
4
2
3
4
2
2
2
)
3 2
(
3
)
4
2
3
2
1
3
2
2
2
[
1,2,4]Triazolo[4,3-a]pyridine analogues 34–43 were accessed via the synthetic routes
presented in Scheme 5. [1,2,4]Triazolo[4,3-a]pyridine-3-carboxylic acids 33a–33e were
produced in two steps via initial condensation of ethyl 2-oxoacetate with 5-substituted 2-
hydrazinylpyridines 32a–32e followed by a PhI(OAc)
2
-mediated ring cyclization. The
carboxylic acids 33a–33e subsequently underwent amide formation with intermediates
1
0 and 31 to afford the desired analogues 34–43.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 112
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
Scheme 5a
0
1
2
3
4
5
6
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
aReagents and conditions: (a) (i) substituted 2-hydrazinylpyridine, ethyl 2-oxoacetate, CH
C, 1 h; (ii) PhI(OAc) , CH Cl , rt, 1 h; (b) (i) LiOH•H O, THF, H O, rt, 20 min; (ii) 2.0 N aqueous HCl;
c) 10, HBTU, i-Pr NEt, DMF, rt, 16 h; (d) 31, HBTU, i-Pr NEt, DMF, rt, 16 h.
OH, 60
3
°
2
2
2
2
2
(
2
2
Acid 44 served as an access point for both piperidine and bicyclic-containing core
analogues 48–53. HBTU-mediated peptide coupling between 44 and amines 10 and 31
furnished compounds 45 and 46, respectively. These intermediary aryl bromides were
converted to the nitrile-bearing analogues 48 and 51 via Pd-catalyzed cyanation with
ZnCN
Analogue 52 was obtained via carbonylation of 46 with Mo(CO)
hydrolysis of the resulting methyl ester with LiOH•H O. Urea 53 was furnished via a Pd-
2
. Methylation of 45 to provide 49 was achieved with Fe(acac)
3
and CH
3
MgBr.
6
in CH
3
OH followed by
2
catalyzed amination between 46 and N,N-dimethylurea. The synthesis of sulfonamide
analogue 54 began with a Pd-catalyzed carbon-sulfur cross-coupling between 46 and
14
ACS Paragon Plus Environment

Page 15 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
benzyl thiol to furnish an intermediary benzyl mercaptan, which was then converted to
the sulfonyl chloride and treated with N-methylamine.
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
Scheme 6a
aReagents and conditions: (a) 10, HBTU, i-Pr
6 h; (c) ZnCN , Pd(PPh , DMF, 130 °C (microwave irradiation), 30 min; (d) Fe(acac)
1.4 M solution in THF/toluene) NMP, THF, rt, 1 h; (e) ZnCN , Pd(PPh , DMF, 130 °C (microwave
irradiation), 30 min; (f) (i) Mo(CO) , Pd(OAc) , xantphos, Cs CO , CH OH, 1,4-dioxane, 80 °C, 2 h;
iii) LiOH•H O, THF, H O, rt, 30 min; (iv) 2 N aqueous HCl; (g) N,N-dimethylurea, Pd(OAc)
xantphos, Cs CO , 1,4-dioxane, 100 °C, 6 h; (h) (i) benzyl thiol, i-Pr NEt, Pd(OAc) , xantphos, 1,4-
dioxane, 110 °C, 16 h; (ii) NCS, HOAC, H O, rt, 3 h; (iii) N-methylamine, i-Pr NEt, CH Cl , rt, 2 h.
2 2
NEt, DMF, rt, 16 h; (b) 31, HBTU, i-Pr NEt, DMF, rt,
1
2
3
)
4
3
, CH MgBr
3
(
2
)
3 4
6
2
2
3
3
(
2
2
2
,
2
3
2
2
2
2
2
2
A subset of analogues presenting alternatively fluorinated 2-trifluoromethyl phenyl
groups were designed to further explore the hydrophobic -ionone binding pocket of
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 112
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
RBP4 (59, 62, 63, 65) (Schemes 7 and 8). The preparation of fluorinated analogue 59 is
depicted in Scheme 7. Conversion of 55 to vinyl triflate 56 was followed by Pd-catalyzed
Suzuki cross-coupling to furnish tetrahydropyridine 57. Hydrogenation of 57 followed by
0
1
2
3
4
5
6
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
N-Boc de-protection with 2.0 M HCl in Et
2
O provided piperidine hydrochloride salt 58.
HBTU-mediated peptide coupling between 58 and carboxylic acid 44 followed by Pd-
catalyzed cyanation with Zn(CN) afforded desired analogue 59.
2
Scheme 7a
aReagents and conditions: (a) (i) LiHMDS (1.0 M solution in THF), THF, -78 °C, 1 h; (ii),
PhN(SO CF , THF, -78 °C to 0 °C, 3 h; (b) (3-fluoro-2-(trifluoromethyl)phenyl)boronic acid,
Pd(PPh , 2.0 M aqueous K CO , DME, 80 °C, 6 h; (c) 10% Pd/C, H (30 psi), EtOH, rt, 18 h; (d) 2.0
M HCl in Et O, CH Cl , rt, 18 h; (e) 44, HBTU, i-Pr NEt, DMF, rt, 16 h; (e) Zn(CN) , Pd(PPh , DMF,
130 °C (microwave irradiation), 30 min.
2
)
3 2
3
)
4
2
3
2
2
2
2
2
2
)
3 4
The preparation of fluorinated analogues 61, 62 and 64 is shown in Scheme 8. Suzuki
cross-coupling between vinyl triflate 56 and (5-fluoro-2-(trifluoromethyl)phenyl)boronic
acid or (4-fluoro-2-(trifluoromethyl)phenyl)boronic acid followed by N-Boc de-protection
gave intermediary tetrahydropyridines 60a and 60b, respectively. PtO -mediated
2
16
ACS Paragon Plus Environment

Page 17 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
hydrogenation of these tetrahydropyridines gave corresponding piperidine intermediates
that readily underwent amide formation with 44. Cyanation of these amides using the
aformentioned Pd-catalyzed conditions with Zn(CN) yielded 61 and 62, respectively. The
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
synthesis of fluorinated aryl head group analogue 64 followed a route previously
described for the preparation of compound 59 with the exception that tetrahydropyridine
hydrogenation to give piperidine intermediate 63 was conducted with PtO
2
under 1 atm
instead of 10% Pd/C at 30 psi of H . N-Boc de-protection of 63 followed by peptide
2
coupling with 44 and subsequent Pd-catalyzed cyanation led to desired compound 64.
Scheme 8a
aReagents and conditions: (a) (i) substituted aryl boronic acid, Pd(PPh
DME, 80 °C, 6 h; (b) 2.0 M HCl in Et O, CH Cl , rt, 18 h; (c) (i) PtO , H (1 atm), HOAc, EtOAc, rt, 72
h; (ii) 2.0 M HCl in Et O, CH Cl , rt, 18 h; (d) 44, HBTU, Et N, i-Pr NEt, rt, 16 h; (e) ZnCN , Pd(PPh
DMF, 130 °C (microwave irradiation), 30 min.
CO ,
3
)
4
, 2.0 M aqueous K
2
3
2
2
2
2
2
2
2
2
3
2
2
3 4
) ,
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 112
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
The syntheses of various 2-carboxamido fused [5,6]-bicyclic heteroaromatic
analogues (65–72) are highlighted in Scheme 9. These analogues were prepared peptide
coupling between the respective heteroaromatic 2-carboxylic acid and 10 in the presence
of HBTU.
0
1
2
3
4
5
6
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
Scheme 9a
aReagents and conditions: (a) 1H-benzo[d]imidazole-2-carboxylic acid, benzo[d]oxazole-2-
carboxylic acid or benzo[d]thiazole-2-carboxylic acid, HBTU, i-Pr
indole-2-carboxylic acid, HBTU, i-Pr NEt, DMF, rt, 16 h; (c) 1H-pyrrolo[2,3-c]pyridine-2-carboxylic
acid, HBTU, i-Pr NEt, DMF, rt, 16 h; (d) 1H-pyrrolo[3,2-b]pyridine-2-carboxylic acid, HBTU, i-
Pr NEt, DMF, rt, 16 h; (e) imidazo[1,2-a]pyridine-2-carboxylic acid, HBTU, i-Pr NEt, DMF, rt, 16 h;
2
NEt, DMF, rt, 16 h; (b) 10, 1H-
2
2
2
2
(
f) imidazo[1,2-b]pyridazine-2-carboxylic acid, HBTU, i-Pr NEt, DMF, rt, 16 h.
2
18
ACS Paragon Plus Environment

Page 19 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The syntheses of a series of follow-up 6-substituted imidazo[1,2-b]pyridazine-2-
carboxamide analogues (74–78) were generated from key intermediate 73 as depicted in
Scheme 10. Methylation of 73 to give 74 was achieved via a Pd(dppf) cross-coupling with
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
trimethylboroxine. Nucleophilic aromatic substitution of 73 with NaOCH
methoxy analogue 75. The installation of the cyclopropyl ring of 76 was achieved via a of
Pd(OAc) cross-coupling reaction with 73 and potassium cyclopropyltrifluoroborate in the
presence di-(1-admantyl)-n-butylphosphine, and Cs CO . Lastly, heating 73 in neat
3
provided
2
2
3
pyrrolidine or morpholine yielded the desired nucleophilic aromatic substitution products
77 and 78, respectively.
Scheme 10a
aReagents and conditions: (a) 6-chloroimidazo[1,2-b]pyridazine-2-carboxylic acid, HBTU, i-Pr
DMF, rt, 16 h; (b) trimethylboroxine, Pd(dppf), K CO , 1,4-dioxane, H O, 110 °C, 5 h; (c) NaOCH ,
2
NEt,
2
3
2
3
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 112
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
CH
3
OH, 70 °C, 1 h; (d) potassium cyclopropyltrifluoroborate, Pd(OAc) , di-(1-admantyl)-n-
2
butylphosphine, Cs
2
CO
3
, toluene, H O, 100 °C, 3 h; (e) pyrrolidine (neat), 100 °C, 3 h; (f)
2
morpholine (neat), 120 °C, 2 h.
0
1
2
3
4
5
6
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 AND DISCUSSION
The reported PDB high-resolution x-ray crystal structures 3fmz (co-crystal structure of 3
bound to RBP4)30, 33 and 4psq (co-crystal structure of 6 bound to RBP4) provided critical
data that was used in support of our structure-based drug design efforts toward novel
fused [5,6]-bicyclic heteroaromatic amide-containing analogues (Figure 4). The structural
data show 3 and 6 binding within the internal binding cavity of RBP4 with similar
geometric alignment. The core piperidine ring of 3 and piperazine ring of 6 both reside in
the relatively narrow -barrel channel of the binding cavity and their respective 2-
trifluoromethyl substituted aryl head group projecting into the vacuous and hydrophobic
-ionone pocket where van der Waals contacts are made (Figure 4). The urea of 3 and the
carboxamide of 6 both accept an H-bond from backbone amide Leu37. A salt bridge forms
between the carboxylic acid of 3 and Arg121 toward the opening of the binding pocket
30
(Figure 4, A). Furthermore, two key H-bonds are also established between the acid of 3,
Tyr90, and Gln98 while the phenyl ring of 3 also engages in an edge-to-face - binding
interaction with Phe96. In contrast, the N2 imidazole nitrogen of 6 engages in an H-bond
with the Arg121 NH group while the N-benzyl appendage projects toward the opening
2
of the binding cavity (Figure 4, B). The key H-bond binding interactions with Leu37 and
Arg121 described for 3 and 6 are hypothesized to play a crucial role in the induction and
20
ACS Paragon Plus Environment

Page 21 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
stabilization of conformational changes within RBP4 loops 3-4 and 5-6 that prohibit
complexation between RBP4 and TTR and afford antagonist functional activity.14, 21, 22, 30
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 4. (A) Binding conformation of anthranilic acid antagonist 3 and key H-bond interactions
with RBP4 as depicted within 3fmz. The image is in stick model format. Carbon atoms are shown
as grey for the ligand and white for the protein, oxygen atoms are red, nitrogen atoms are blue,
chlorine atoms are green, and sulfur atoms are yellow. H-bonds are indicated as blue dotted lines
and the residues undergoing H-bond interactions with 3 are labeled. The urea carbonyl of 3
engages in an H-bond interaction with backbone Leu37 while Arg121 forms a salt bridge and Glu98
H-bonds with the carboxylic acid. In addition, the anthranilic acid phenyl ring of 3 and Phe96 are
engaged in an edge-to-face - binding interaction (not shown). Lastly, the distal 2-
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 112
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
trifluoromethyl substituted aromatic head group of 3 resides with the hydrophobic and vacuous

-ionone pocket and is positioned nearly orthogonal to the core piperidine ring. The image is of
PDB ID 3fmz was obtained from the RCSB Protein Data Bank (PDB) (www.rcsb.org). (B) Bioactive
conformation and binding interactions of N-benzyl imidazole antagonist 6 with RBP4 as depicted
within 4psq. The amide carbonyl of 6 engages in an H-bond interaction with Leu37 and 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
imidazole N2 nitrogen H-bonds with the NH group of the Arg121 side chain. The distal 2-Cl
2
substituted aromatic head group of 3 also resides with the hydrophobic -ionone pocket in a
nearly orthogonal position. The image is of PDB ID 4psq was obtained from the RCSB Protein Data
Bank (PDB) (www.rcsb.org).
As previously reported, 3fmz data was employed to construct a computational
docking model (version 5.8, Schrödinger, LLC.), which was used to triage novel analogue
proposals prior to synthesis by identifying compounds that were predicted to have good
docking orientations within the RBP4 binding cavity. The docking and analysis of novel
analogues in our 3fmz data model of RBP4 was performed as previously described.26, 27
Drawing from the N-benzyl imidazole motif of 6, we hypothesized that analogues
presenting carboxamides of fused [5,6]-bicyclic heteroaromatic systems could also serve
to position key H-bond accepting groups within close proximity to Arg121 and Leu37.
Furthermore, such motifs remove the rotatable bonds of the N-benzyl appendage of 6
but still project and aromatic ring toward the opening of the binding pocket, which may
provide analogues with improved potency and physicochemical properties. Indeed, our
initial 3fmz docking experiment with 3-carboxamido-1H-indazole 11 predicted that the
N2 nitrogen of the 1H-indazole ring H-bonds with Arg121 while the carboxamide carbonyl
22
ACS Paragon Plus Environment

Page 23 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
oxygen would serve as an H-bond acceptor with Leu37 (Figure 5). Encouraged by the
docking result for 11, we synthesized a sample set of 1H-indazole- (11–21) and
benzisoxazole- (22 and 23) containing analogues.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Comparison of the putative key binding interactions between anthranilic acid 3 and 3-
carboxamido 1H-indazole 11 with RBP4. (A) Expanded view of the minimized bound conformation
of 3. H-bonds with Gln98 and Tyr90 and a salt bridge with Arg121 are shown (blue lines). An
additional H-bond between the urea carbonyl oxygen of 11 with the backbone amide Leu37 is
also present but out of view in this depiction. Hydrophobic molecular surfaces are shown as green
and hydrophilic surfaces as purple. Contact preferences generated by MOE (Chemical Computing
Group, Inc., Montreal, CA). (B) Expanded view of the 1H-indazole fragment of 11. The 1H-indazole
N2 nitrogen atom accepts an H-bond from Arg121. An additional H-bond between the amide
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 112
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
carbonyl oxygen of 11 with the backbone amide Leu37 is also present but out of view in this
depiction. Furthermore, a potential van der Waals interaction may be occurring between the 1H-
indazole ring and the isobutyl group of Leu37. The carbon atoms of 11 are grey and the carbon
atoms of the resident RBP4 amino acid residues are green. Nitrogen atoms are shown as blue and
oxygen atoms are shown as red. Grey lines indicate van der Waals contacts.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In Vitro Binding of Compounds to RBP4. Binding affinities for RBP4 were measured
using our scintillation proximity assay (SPA) as previously reported.26-28, 34 Binding
experiments were performed on white 96-well plates in a final assay volume of 100
3
μl/well. The reaction mixture contained 10 nM H-Retinol (48.7Ci/mmol; PerkinElmer),
0
.3 mg/well Streptavidin-PVT beads, 50 nM of RBP4 labeled with biotin. The SPA buffer
contained 1X PBS, pH=7.4, 1 mM EDTA, 0.1% BSA, and 0.5% CHAPS. Nonspecific binding
was measured in the presence of 20 μM of cold (unlabeled) retinol. Using these assay
conditions, saturation binding of retinol to RBP4 afforded a K
d
of 62.5 nM, which is in line
with the previously reported 70–190 nM range of values. For compound characterization,
3
3
competitive displacement of H-retinol was measured. The concentration of H-retinol
was 10 nM, which corresponds to the generally recommended concentration of
radioligand (1/10 of the K ). The IC50 values were calculated from 8-point compound
d
titrations using four-parameter nonlinear regression in GraphPad Prism (Version 8).
Compounds with appreciable RBP4 binding potency (generally, with IC50 <150 nM) were
further assessed for their ability to inhibit the retinol-dependent RBP4-TTR interaction
using a homogenous time resolved fluorescence (HTRF) assay. 26-28, 34
24
ACS Paragon Plus Environment

Page 25 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Antagonism of RBP4-TTR Interaction HTRF Assay. The ability of the compounds to
disrupt the retinol-induced interaction of RBP4 with TTR was examined using our HTRF
assay as described previously.26-28, 34 Bacterially expressed Maltose Binding Protein
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
+
(
MBP)-tagged RBP4 and Eu -cryptate-labeled commercially available human TTR were
used along with a d2-conjugated anti-MBP monoclonal antibody. Retinol at 1 µM
concentration stimulates the formation of the RBP4-TTR complex, which brings europium
in close proximity to the d2 dye, resulting in the fluorescence energy transfer (FRET) to
d2. The FRET emission signal (668 nm) was normalized using the europium emission (620
nm) to compensate for the pipetting and dispensing errors. Following a 12-point dose
titration (30 µM–0.1 nM), the IC50 values were calculated using four-parameter nonlinear
regression in GraphPad Prism (Version 8). To monitor the correlation between RBP4
binding affinity of the compounds and their potency as antagonists of the retinol-
dependent RBP4-TTR interaction, the ratio of their IC50 values in the HTRF and SPA assays
was tracked. For most compounds, a ratio in the 5–20 range was observed, despite the
differences in their binding affinities.
Analogue SAR. 1H-indazole-3-carboxamide analogues containing substituents at the
5
-, 6-, and 7-position of the 1H-indazole ring (11, 13–18) and on the N1 nitrogen atom
(12, 19–21) were initially prepared from piperidine hydrochloride 10. In vitro RBP4
potency, kinetic aqueous solubility, and liver microsomal metabolic stability for these
compounds are shown in Table 1. The majority of these analogues exhibited excellent in
vitro potency in the RBP4 SPA and HTRF assays. 1H-Indazole 11 presented an approximate
2
-fold improvement in RBP4 SPA binding affinity relative to benchmark 3 and a nearly 3-
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 112
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
fold improvement in HTRF functional antagonist activity. Substitution with either F or Cl
at the 5-, 6- or 7-position of the 1H-indazole ring (13–16, 18) provided analogues of
comparable potency to parent 11. However, installation of an OCH
position (17) significantly diminished RBP4 SPA and HTRF potency. It is unclear whether
the increased steric bulk, polarity, or the electron-donating capability of the OCH group
or any combination thereof) contributed to the loss of potency observed for 17 relative
to 13–16 and 18.
3
group at the 5-
0
1
2
3
4
5
6
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
(
The N1 1H-indazole nitrogen also provided a handle with which to install various
groups that could further probe the RBP4 binding cavity. Methylation to give 12 neither
improved nor diminished in vitro RBP4 potency relative to parent 11, whereas increasing
the bulk of the N1 substituent from methyl to ethyl (19) and isopropyl (20) significantly
reduced potency. Interestingly, installation of the N1 oxetane appendage (21) restored
potency comparable to 12. Lastly, isosteric benzisoxazoles 22 and 23 were also found to
be of comparable potency to 11. The favorable in vitro RBP4 potencies observed for this
initial sample set confirmed that fused [5,6]-bicyclic systems may be viable replacements
for the carboxylic of 3 and 5 and the N-benzyl imidazole of 6. However, enthusiasm for
this series was tempered as the compounds generally exhibited very poor kinetic aqueous
solubility or poor microsomal metabolic stability (or both), which was likely attributable
in part to their relatively high lipophilicity and sub-optimal lipophilic ligand efficiency (LLE)
(
RBP4 SPA pIC50 - cLogP).35
26
ACS Paragon Plus Environment

Page 27 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. RBP4 Binding Affinity, Functional RBP4-TTR Antagonism, Kinetic Aqueous
Solubility, and Liver Microsomal Metabolic Stability Data for 3-Carboxamido Indazole
and Benzisoxazole RBP4 Antagonists.
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
RBP4
HTRFa
Kinetic
Solubilityb
(M)
Microsomal Stability
(% remaining)
Compound
Number
RBP4 SPAa
IC50 (nM)
R
cLogP
LLE
HLMc RLMd MLMe
IC50 (nM)
1
5.0 ±
3
122± 0.035
ND
3.0
85
ND
4.85
2.97
0.005
11
6.42
46
7.4
3.6
3
7.5
0.7
68
0.0
0.0
0.2
0.6
0.2
8.7
4.21
3.87
4.41
3.98
4.56
3.91
1
1
2
3
3.71
4.71
56.1
49.0
1
1
4
5
7.53
4.71
67.1
48.3
< 1.6
1.9
100
36
0.1
0.5
100
6.8
4.98
4.41
3.14
3.91
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 112
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
6
7
6.99
44.6
58.6
843
< 1.6
ND
100
ND
0.8
ND
51
4.98
4.34
3.17
3.01
ND
0
1
2
3
4
5
6
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
18
5.47
9.54
128
82.5
242
ND
2.5
2.1
ND
78
1.3
ND
0.2
0.3
ND
34
1.2
ND
4.98
4.55
4.86
3.28
3.47
2.03
1
2
9
0
21
4.12
35.6
4.7
0.1
10
4.3
3.57
4.81
2
2
2
3
6.28
8.26
79.5
98.8
8.4
6.2
0.4
1.5
0.0
0.0
0.0
0.0
3.97
3.97
4.23
4.11
aFor compounds tested more than twice, IC50 is represented as the mean ± standard deviation.
b
Otherwise, IC50 is shown as the mean of two independent experiments. Kinetic aqueous solubility
measured in PBS, pH 7.4. Verapamil and tamoxifen were used as controls. cHLM = human liver
d
e
microsomes. RLM = rat liver microsomes. MLM = mouse liver microsomes. Compound
concentration was 10 M and incubation time with either human, rat or mouse microsomes was
28
ACS Paragon Plus Environment

Page 29 of 112
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
0 minutes. Testosterone was used as a positive control. cLogP derived from ChemDraw
Professional (ChemOffice Professional, CambridgeSoft and Perkin Elmer). LLE = lipophilic ligand
efficiency: RBP4 SPA pIC50 – cLogP. ND = not determined.
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
We next explored less lipophilic [1,2,4]triazolo[4,3-a]pyridine-3-carboxamides with
the piperidine core of 10 (34–38, 48, 49) and the bicyclic core of 31 (39–43, 50–53). Data
for these compounds are highlighted in Table 2. We were encouraged to find that the
[
1,2,4]triazolo[4,3-a]pyridine motif afforded antagonists with excellent in vitro potency.
Importantly, a direct comparison of the fluorinated 1H-indazole 15 with its less lipophilic
congener 34 revealed that the latter exhibits comparable SPA potency, improved HTRF
potency, and improved LLE (15 LLE = 4.41; 34 LLE = 5.45). Indeed, the [1,2,4]triazolo[4,3-
a]pyridine series as a whole demonstrated a general improvement in LLE trends relative
to the previous 1H-indazole series, which also correlated with improved solubility and
metabolic stability for compounds within this sample set, as exemplified by nitrile 48 (LLE
=
6.08).
29
ACS Paragon Plus Environment