Identification of potent pyrazole based APELIN receptor (APJ) agonists

Article abstract of DOI:10.1016/j.bmc.2019.115237

The apelinergic system comprises the apelin receptor and its cognate apelin and elabela peptide ligands of various lengths. This system has become an increasingly attractive target for pulmonary and cardiometabolic diseases. Small molecule regulators of this receptor with good drug-like properties are needed. Recently, we discovered a novel pyrazole based small molecule agonist 8 of the apelin receptor (EC₅₀ = 21.5 μM, K_i = 5.2 μM) through focused screening which was further optimized to initial lead 9 (EC₅₀ = 0.800 μM, K_i = 1.3 μM). In our efforts to synthesize more potent agonists and to explore the structural features important for apelin receptor agonism, we carried out structural modifications at N1 of the pyrazole core as well as the amino acid side-chain of 9. Systematic modifications at these two positions provided potent small molecule agonists exhibiting EC₅₀ values of <100 nM. Recruitment of β-arrestin as a measure of desensitization potential of select compounds was also investigated. Functional selectivity was a feature of several compounds with a bias towards calcium mobilization over β-arrestin recruitment. These compounds may be suitable as tools for in vivo studies of apelin receptor function.

Full text of DOI:10.1016/j.bmc.2019.115237

Journal Pre-proofs
Identification of Potent Pyrazole Based APELIN Receptor (APJ) Agonists
Sanju Narayanan, Vineetha Vasukuttan, Sudarshan Rajagopal, Rangan
Maitra, Scott P. Runyon
PII:
S0968-0896(19)31551-2
https://doi.org/10.1016/j.bmc.2019.115237
BMC 115237
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
9 September 2019
19 November 2019
23 November 2019
Please cite this article as: S. Narayanan, V. Vasukuttan, S. Rajagopal, R. Maitra, S.P. Runyon, Identification of
Potent Pyrazole Based APELIN Receptor (APJ) Agonists, Bioorganic & Medicinal Chemistry (2019), doi:
https://doi.org/10.1016/j.bmc.2019.115237
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2019 Published by Elsevier Ltd.

Identification of Potent Pyrazole Based APELIN
Receptor (APJ) Agonists
Sanju Narayanan†, Vineetha Vasukuttan†, Sudarshan Rajagopal‡, Rangan Maitra†, and Scott P. Runyon*†
†
Center for Drug Discovery, RTI International, Research Triangle Park, NC 27709, United States, ‡Center for Pulmonary Vascular Disease,
Duke University Medical Center, Durham, NC 27710, United States
Keywords: Apelin, APJ, agonist, pyrazole, AGTRL1, APLNR

Abstract
The apelinergic system comprises the apelin receptor and its cognate apelin and elabela peptide
ligands of various lengths. This system has become an increasingly attractive target for pulmonary
and cardiometabolic diseases. Small molecule regulators of this receptor with good drug-like
properties are needed. Recently, we discovered a novel pyrazole based small molecule agonist 8
of APJ (EC50 = 21.5 µM, K
i
= 5.2 µM) through focused screening which was further optimized to
= 1.3 µM). In our efforts to synthesize more potent agonists
initial lead 9 (EC50 = 0.800 µM, K
i
and to explore the structural features important for apelin receptor agonism, we carried out
structural modifications at N1 of the pyrazole core as well as the amino acid side-chain of 9.
Systematic modifications at these two positions provided potent small molecule agonists
exhibiting EC50 values of < 100 nM. Recruitment of -arrestin as a measure of desensitization
potential of select compounds was also investigated. Functional selectivity was a feature of several
compounds with a bias towards calcium mobilization over-arrestin recruitment. These
compounds may be suitable as tools for in vivo studies of APJ function.
2

Introduction
The apelinergic system comprises the endogenous apelin peptides and its cognate G-protein
1
coupled receptor, APJ, encoded by the gene AGTRL1 (angiotensin receptor like 1). The apelin
receptor is considered physiologically important due to its presence across many species including
2
human, monkey, pig, rodents, frog and zebrafish. The apelin receptor is a 380 amino acid protein
1
having ~33% sequence identity to the AT1 receptor. Apelin-36 was the first endogenous peptide
identified along with other smaller peptides (apelin-17, apelin-13, apelin-12, pyr-apelin-13) that
were derived from the cleavage of paired basic residues (Arg-Arg and Arg-Lys) of the 77 amino
3
-7
8
acid precursor protein. Known apelin peptides are readily hydrolyzed in vivo (half-life ~5 min).
These proteins belong to an understudied component of the renin-angiotensin system with
angiotensin-converting-enzyme II (ACE II) being involved in degradation of apelin in vivo.8,9
However, none of the known angiotensin receptor ligands or peptides activate apelin receptor.1
Recently, Elabela/Toddler was identified as the second endogenous ligand of this receptor.10-12
Identification of these two sets of endogenous ligands facilitated molecular characterization of
apelin receptor. Apelin receptors are primarily coupled to inhibitory Gαi proteins and their
3
activation is associated with inhibition of forskolin-stimulated cAMP production. Additionally,
apelin receptor can also activate PLC mediated calcium flux through Gαq coupling13,14 and recruit
15
β-arrestin that desensitizes the receptor utilizing clathrin mediated endocytosis. Apelin and apelin
receptor are widely expressed across tissues in both human and rodents. In rat CNS, apelin mRNA
is expressed within the hypothalamus where it is co-expressed with vasopressin mRNA in
paraventricular (PVN) and supraoptic nuclei (SON).16,17 Peripherally, apelin receptor is expressed
in many organs including kidney, lungs, adipose tissue, liver and heart.18-20 Activation of the
apelinergic system has many beneficial effects including protection against hypertensive disorders,
3

excitotoxic neuronal damage, pulmonary arterial hypertension, metabolic syndrome and heart
failure.21,22 The receptor also plays an important role during embryonal development through
regulation of vasculogenesis.23
Efforts are underway to produce small molecule agonists of apelin receptor with improved
drug-like properties because the endogenous peptide ligands have short half-lives and limited
utility as therapeutics. A number of small molecules have been reported. Iturrioz and colleagues
reported a non-peptide functional apelin receptor agonist E339-3D6 (Figure 1) that exhibited an
EC50 of 90 nM (FRET) and K
i
value of ~400 nM in both rat and human apelin receptor.24
Figure 1. Non-peptide apelin receptor agonists reported in the literature
The
ML233
core
(Figure1)
((E)-2-cyclohexyl-5-methyl-4-
(phenylsulfonyloxyimino)cyclohexa-2,5-dienone) was identified through a high throughput screen
of ~330 600 compounds from the NIH small molecule library.25 ML233 is a full agonist and
exhibits an EC50 of 3.7 µM (β-arrestin), but it displayed poor solubility and stability in microsomal
4

fractions. In addition, ML233 had significant off-target binding at several other receptors
including the 5-HT1A, α2C adrenergic, mu-opioid, benzylpiperazine receptors and norepinephrine
2
6
transporter. A number of patent disclosures have also emerged. Sanofi Aventis, and Sanford-
Burnham Medical Research Institute,27 have disclosed small molecule agonists based on the
substituted benzimidazole (US20140094450 A1) and triazole core (WO 2015/184011 A2)
respectively (Figure 1). In addition, Bristol-Myers Squibb and Amgen have disclosed small
2
8
molecules based on dihydropyrimidine-4-one core (WO 2017/096130A1) and triazole core (WO
2
9
2
017/192485A1) respectively. A small molecule benzimidazole derivative related to the Sanofi
Aventis scaffold CMF-019 (Figure 1) was reported to be G protein biased small molecule apelin
receptor agonist. CMF-019 was ~400 fold bias towards Gαi signaling over β-arrestin recruitment
and ~6000 fold bias over receptor internalization. Studies using these biased agonists that
preferentially stimulate G-protein activation over β-arrestin recruitment have demonstrated
improved in vivo efficacy (vasodilation and inotropic actions) without β-arrestin dependent
cardiac hypertrophy.30-32
Our laboratory has previously reported a novel pyrazole scaffold 8 (EC50 = 21.5 µM
(
calcium), K
i
= 5.2 µM) identified through focused screening, which was further optimized to an
= 1.3 µM) (Figure 2).33 In our efforts to identify
initial lead 9 (EC50 = 0.800 µM (calcium), K
i
more potent agonists as tool compounds and to explore the structural features important for apelin
receptor agonism, we carried out modifications at N1 position of the pyrazole core as well as the
amino acid side-chain of the lead compound 9. Synthesized compounds were characterized using
calcium mobilization, inhibition of forskolin-induced cAMP accumulation and β- arrestin
recruitment assays. Considering the potential importance of selectivity over the -arrestin
3
0,31
pathway, bias factors for the most promising compounds were calculated.
5

Figure 2. Chemical structure of initial hit 8, identified through focused screen and optimized to
lead compound 9
Chemistry
Scheme 1 depicts the synthetic route employed to prepare target compounds 9 – 38. Synthesis of
analogs 25 and 37 are shown as representative examples. 2, 6-Dimethoxy acetophenone 39 was
condensed with diethyl oxalate to afford the sodium salt of diketone 40 in quantitative yield.
Reaction of 40 with isobutylhydrazine or cyclopentylhydrazine trifluoroacetate in refluxing
ethanol provided 1, 5-pyrazoles 41a, 42a and 1,3-pyrazoles 41b, 42b in a ratio of 4:1.
Transformation of 41a and 42a to acids 43 and 44 was achieved using alkaline hydrolysis with
lithium hydroxide monohydrate. Coupling of carboxylic acids 43 and 44 with (S)-tert-butyl 3-
amino-5-cyclohexylpentanoate33,34
in
the
presence
of
benzotriazole-1-yl-oxy-tris-
(dimethylamino)-phosphonium hexafluorophosphate (BOP) afforded 45 and 46. Tert-butyl esters
4
5 and 46 were deprotected using trifluoroacetic acid to provide acids 12 and 13. Finally,
carboxylic acids 12 and 13 were coupled to cyclobutylamine using BOP to obtain 25 and 37
respectively.
6

Scheme 1: Synthesis of compound 25 and 37
Scheme 1: Reagents and conditions: a) Diethyl oxalate, EtONa, EtOH, reflux, 6 h, 97%; b)
Isobutylhydrazine trifluoroacetate/cyclopentylhydrazine trifluoroacetate, EtOH, reflux, 16 h;
4
0/73% c) LiOH, MeOH/THF/H
cyclohexylpentanoate, BOP, Et N, THF, rt, 1.5 h, 91/87%; e) TFA, DCM, rt, 1.5 h, 54/94%; f)
cyclobutylamine, BOP, Et N, THF, rt, 3 h, 61/75%.
2
O, rt, 18 h, 96/98%; d) (S)-tert-butyl 3-amino-5-
3
3
As shown in Scheme 1 the synthesis of compounds 41 and 42 provides a mixture of isomers at
the N1 and N2 pyrazole positions in ratio 4:1 that are readily separable using normal phase
chromatography. The structures of hydrolyzed carboxylic acids 43 and 44 were not previously
described but were confirmed using X-Ray crystallography (Figures 3 & 4). The NMR shift
7

assignments for 41 and 42 were used as a guide to identify the additional N1 vs N2 alkyl substituted
pyrazoles synthesized in this study.
Figure 3. Crystal structure of compound 43
Figure 4. Crystal structure of compound 44
8

Results and Discussion
Our laboratory has previously reported the novel pyrazole based small molecule agonist
2
+
hit 8 (Ca EC50 = 21.5 µM, K
optimized to an initial lead 9 (Ca EC50 = 0.800 µM, cAMP EC50 = 844 nM, K
addition to improving potency and affinity, compound 9 exhibited significantly reduced off-target
i
= 5.2 µM) identified through focused screening that was further
2
+
i
= 1.3 µM). In
3
3
NTR2 activity and did not activate or inhibit the AT1 receptor (Table 1). Considering the
structural similarity with the Sanofi scaffold 7, we hypothesized that modification of 9 at N1
position could potentially enhance apelin receptor potency. Substitution of 4-fluorophenyl in 9
with various alkyl or cycloalkyl groups were initially considered as these were envisioned to
provide similar hydrophobicity and steric bulk (Table 1) as phenyl. In addition, these analogs
would also help determine whether aromatic substitutions were optimal at the N1 position for
apelin receptor potency with this scaffold. The agonist activities of synthesized analogs were
evaluated using both calcium mobilization and cAMP generation functional assays. In addition to
potency and efficacy, recruitment of -arrestin was also evaluated.
9

Table 1. Modification at N1 pyrazole core of compound with cyclohexyl ethyl amino acid side
chain
APJ
APJ
APJ
APJ
AT1
Calcium
cAMP
β-arrestin
Binding
Calcium
Compd.
R
EC50
*
EC50
*
EC50
*
K
i
*
EC50*
µM±SEM
µM±SEM
µM±SEM
(µM±SEM) (µM±SEM)
(
%Emax
)
(%Emax
)
(%Emax)
Pyr-Apelin-13
-
0.002 ± 0.0003
0.800 ± 0.1 (91)
1.46 ± 0.4 (65)
0.0003 ± 0.0001
0.844 ± 0.2 (103)
2. 55 ± 1.6 (106)
0.001 ± 0.0005
6.22 ± 2.0 (86)
9.94± 1.4 (75)
0.001 ± 0.0005
ND*
>10
ND
ND
ND
ND
ND
9
.
1.3 ± 0.3
ND
1
1
1
1
1
0.
1.
2.
3.
4.
CH
3
CH
2
-CH
2
-CH
3
0.416 ± 0.02 (89) 0.281 ± 0.09 (109) 3.28 ± 0.4 (67)
0.649 ± 0.1 (101) 0.915 ± 0.5 (108) 5.59 ± 0.06 (81)
0.424 ± 0.07 (96) 0.238 ± 0.05 (103) 2.47 ± 0.4 (77)
0.380 ± 0.03 (91) 0.518 ± 0.08 (113) 3.26 ± 0.08 (83)
ND
ND
ND
ND
1
1
5.
6.
>10
>10
>10
>10
>10
>10
ND
ND
ND
ND
*
*
i
EC50 and K values are averages of multiple experiments performed in duplicate unless otherwise stated ± standard
error of the mean.
ND = Not Done
A. Modification at N1 pyrazole core of compound with cyclohexyl ethyl amino acid side chain:
2
+
The N1 methyl substituted analog 10 (Ca EC50 = 1.46 µM, cAMP EC50 = 2.55 µM) exhibited
1
0

2
+
approximately 2-fold reduced potency compared to the aromatic N1 phenyl substituent 9 (Ca EC50
=
0.800 µM). Increasing the chain length to three carbon propyl 11, enhanced potency by 2-fold
2
+
(
Ca EC50 = 0.416 µM) compared to 9. Adding bulk with addition of branched alkyl substitution
2
+
(
isobutyl) 12, however, did not enhance potency (Ca EC50 = 0.649 µM). Further, increasing the
2+
hydrophobic volume to five carbon cyclopentyl 13 (Ca EC50 = 0.424 µM) and six carbon
cyclohexyl 14 (Ca2+ EC50 = 0.380 µM) retained the 2-fold enhanced potency. Extending the
cyclohexyl ring by one methylene 15 (Ca2+ EC50 >10 µM) significantly reduced the potency
suggesting tolerance for alkyl/cycloalkyl moiety with optimum hydrophobic volume of 3-6 carbons.
This was further confirmed by replacing 4-fluorophenyl with larger hydrophobic cyclooctyl moiety
1
6 (Ca2+ EC50 >10 µM) that proved detrimental for activity. The agonist activities of these
compounds were also evaluated using cAMP assay to monitor the influence of these modifications
on functional selectivity. Compounds 11 (cAMP EC50 = 0.281 µM), 13 (cAMP EC50 = 0.238 µM)
and 14 (cAMP EC50 = 0.518 µM) exhibited 2 to 3-fold enhanced potency compared to the N1
phenyl substituted lead compound 9 (cAMP EC50 = 0.844 µM). By contrast, 12 had similar potency
to 9 in the cAMP assay. Compounds 15 and 16 with bulky cyclohexyl methyl and cyclooctyl group
were completely inactive. Recruitment of β-arrestin was also investigated as a potential measure of
desensitization. Compounds 11, 12, 13, 14 exhibited > 2 µM potencies in -arrestin assay
suggesting the possibility of reduced receptor internalization and desensitization with these analogs.
These studies suggested that an aryl ring was not required at N1 position of the pyrazole core to
retain APJ receptor activity. Compounds substituted with alkyl/cycloalkyl moieties and with
optimum hydrophobic volumes (n-propyl, cyclopentyl, cyclohexyl) exhibited enhanced potency
and preference towards calcium/cAMP signaling over β-arrestin recruitment.
1
1

Table 2. Modification of carboxylic acid moiety of 12
APJ
APJ
APJ
APJ
AT1
Calcium
cAMP
β-arrestin
Binding
Calcium
Compd.
R
EC50
*
EC50
*
EC50
*
K
i
*
EC50*
µM±SEM
µM±SEM
µM±SEM
(µM±SEM) (µM±SEM)
(
%Emax
)
(%Emax
)
(%Emax)
1
1
1
2
2
2
7.
8.
9.
0.
1.
2.
-NH
2
5.75 ± 1.1 (94)
0.288 ± 0.03 (92)
0.237 ± 0.04 (91)
1.85 ± 0.5 (77)
>10
7.24 ± 0.9 (133)
0.280 ± 0.001 (89)
0.453 ± 0.014 (87)
8.87 ± 2.7 (87)
ND*
ND
ND
ND
ND
ND
>10
>10
0.684 ± 0.1 (105)
1.85 ± 0.2 (106)
4.82 ± 1.1 (129)
ND
ND
0.046 ± 0.008 (88) 0.118 ± 0.02 (104)
0.113 ± 0.02 (89)
0.036 ± 0.006
0.038 ± 0.01
0.069 ± 0.02 (139) 0.157 ± 0.08 (104) 0.114 ± 0.005 (90)
1
2

2
2
2
3.
4.
5.
0.144 ± 0.03 (102) 0.246 ± 0.02 (106) 0.182 ± 0.045 (87)
0.172 ± 0.1
>10
ND
>10
1.70 ± 0.04 (94)
0.081 ± 0.02 (91)
3.17 ± 0.8 (134)
0.283 ± 0.08 (97)
0.313 ± 0.04 (81)
0.144 ± 0.008 (90)
ND
0.054 ± 0.01
*
*
i
EC50 and K values are averages of multiple experiments performed in duplicate unless otherwise stated ± standard
error of the mean.
ND = Not Done
B. Modification of carboxylic acid moiety of 12: Next, we focused on optimizing the carboxyl acid
of 12. Simple conversion of carboxylic acid to primary amide 17 (Table 2) was detrimental for
activity (Ca2+ EC50 > 5.75 µM). This suggested that either an acidic group or bulky amide
substitution was necessary to retain activity. The first substituted amide compound 18 with glycine
methyl ester displayed ~2-fold enhanced potency (Ca2+ EC50 = 0.288 µM) compared to the
2
+
carboxylic acid derivative 12 (Ca EC50 = 0.649 µM) suggesting substituted amides versus acids
could also retain or enhance potency. The N-methyl analog 19 was synthesized to probe the
importance of amide proton for receptor interaction. Compound 19 exhibited similar potency (Ca2+
EC50 = 0.237 µM) compared to the unsubstituted amide, suggesting the amide proton is not critical
for activity through the calcium signaling pathway. In contrast, compound 19 exhibited 3-fold
reduced potency (cAMP EC50 = 1.85 µM) in cAMP assay confirming the importance of the amide
2
+
N-H to engage this pathway. Similarly, conversion of the glycine ester 18 (Ca EC50 = 0.288 µM)
2
+
to glycine acid 20 displayed 6-fold reduced potency (Ca EC50 = 1.85 µM) signifying a preference
for the hydrophobic ester over the polar carboxylic acid group. Since esters are metabolically labile
and easily hydrolyzed in vivo, we synthesized relatively stable analogs 21 and 23 with 2-butanol
1
3

and 2-butanone substitution respectively. Interestingly, compound 21 and 23 displayed 14-fold
2
+
2+
(
Ca EC50 = 0.046 µM) and 4-fold (Ca EC50 = 0.144 µM) enhanced potency compared to the
2
+
carboxylic acid derivative 12 (Ca EC50 = 0.649 µM). In addition, compound 25 with cyclobutyl
2
+
amide substitution also exhibited ~ 8-fold enhanced potency (Ca EC50 = 81 nM). Taken together,
these data suggest the hydrophobic amide substitution is important to enhance agonist potency.
However, replacing glycine methyl ester with the more stable ethyl ether 24 exhibited reduced
2
+
potency (Ca EC50 = 1.70 µM). Compounds 17-25 were also evaluated for their ability to inhibit
forskolin stimulated cAMP accumulation. Compounds 21 (cAMP EC50 = 0.118 µM), 22 (cAMP
EC50 = 0.157 µM), 23 (cAMP EC50 = 0.246 µM) and 25 (cAMP EC50 = 0.283 µM) exhibited 2 to
3
-fold enhanced potency compared to the carboxylic acid analog 12 (cAMP EC50 = 0.915 µM).
Compounds 18, 21, 22, 23, 24, 25 were also potent in -arrestin recruitment and exhibited an EC50
value < 0.5 µM. Thus, conversion of carboxylic acid to alkyl and cycloalkyl amides retained or
enhanced potency for -arrestin recruitment indicating a preference for hydrophobic amides is
favorable for activating this signaling pathway. Binding affinities of select compounds were also
1
25
evaluated using [ I]-apelin-13 in radioligand displacement assay. High binding affinities of
compounds 21 (K = 0.036 µM), 22 (K = 0.038 µM), 23 (K = 0.172 µM), 25 (K = 0.054 µM)
i
i
i
i
1
25
indicated that these compounds competitively displaced [ I]-apelin-13 and were bound to the
orthosteric site of apelin receptor (Table 2). Compounds 21, 22, 23, 25 also did not activate AT1
receptor (EC50 > 10 µM), exhibiting selectivity of cyclohexyl ethyl series over the off-target AT1
receptor.
1
4

Table 3. Modification of carboxylic acid moiety of 26.
APJ
Calcium
EC50*
APJ
cAMP
EC50*
APJ
β-arrestin
EC50*
APJ
Binding
AT1
Calcium
EC50 *
n
Compd.
R
K
i
*
µM±SEM
µM±SEM
µM±SEM
(µM±SEM) (µM±SEM)
(
%Emax
)
(%Emax
)
(%Emax
)
(%Emax
)
2
2
6.
7.
-OH
0
0
>10
>10
>10
ND*
ND
ND
0.021± 0.007 (131)
3.09 ± 3.0 (124)
2.33 ± 0.9 (119)
0.187± 0.06
2
2
3
8.
9.
0.
0
0
0
0.027 ± 0.02 (106)
0.411± 0.09 (97)
0.313± 0.01 (93)
0.266 ± 0.057 (106) 0.207 ± 0.002 (99)
0.134± 0.09
ND
>10
ND
ND
>10
>10
>10
>10
ND
3
3
1.
2.
0
0
0.162± 0.001 (98)
1.26 ± 0.8 (81)
>10
8.67± 0.1 (54)
>10
0.571± 0.3
ND
>10
ND
3.48± 0.1 (13)
3
3
3.
4.
-OH
1
1
>10
>10
>10
>10
ND
ND
>10
0.083± 0.01 (96)
3.11 ± 0.2 (80)
0.449± 0.08
*
i
EC50 and K values are averages of multiple experiments performed in duplicate ± standard error of the mean.
*
ND = Not Done
1
5

C. Modification of carboxylic acid moiety of 26: Early on in our program we had evaluated if
phenyl ethyl could replace the cyclohexyl ethyl sidechain because structural modifications of an
aryl ring appeared more synthetically feasible. Since we had identified amide substituents that
greatly enhanced agonist potency such as cyclobutyl amide (25), we evaluated if potency for (26,
2
+
Ca EC50 >10 µM) could be rescued with a similar set of modifications of the carboxylic acid
moiety as shown in Table 2. All of the new amides with the phenylethyl sidechain (Table 3) had
significantly enhanced potency over the parent acid 26, in keeping with the observations for the
cyclohexyl ethyl analogs in Table 2. Compound 27 with glycine methyl ester exhibited > 400-
2
+
fold (Ca EC50 = 0.021 µM) enhanced potency compared to the carboxylic acid derivative 26
2
+
(
Ca EC50 >10 µM). In addition, compound 27 displayed > 100-fold functional selectivity toward
calcium signaling over cAMP (cAMP EC50 = 3.09 µM) and β-arrestin pathway (β-arr EC50 = 2.33
µM). However, functional selectivity was lost when the amide proton was replaced with methyl in
2
+
2+
analog 28 (Ca EC50 = 0.027 µM, cAMP EC50 = 0.266 µM). Similarly, analogs 29 (Ca EC50
=
2
+
0
.411 µM, cAMP EC50 >10 µM, β-arr EC50 >10 µM), 30 (Ca EC50 = 0.313 µM, cAMP EC50 >10
2
+
µM, β-arr EC50 >10 µM), 31 (Ca EC50 = 0.162 µM, cAMP EC50 >10 µM, β-arr EC50 = 8.67 µM)
with ethyl ether, 2-butanol and 2-butanone substitution were also selectively potent in the calcium
2
+
assay. The cyclobutyl amide compound 32, however, had moderate potency (Ca EC50 = 1.26
µM, cAMP EC50 > 3.48 µM) and did not demonstrate signaling bias. Interestingly, the cyclobutyl
2
+
amide β-amino acid analog 34 displayed enhanced potency (Ca EC50 = 0.083 µM) and was also
selective towards calcium signaling over cAMP pathway. Thus, elongation of the carboxylic acid
by one methylene in 34 demonstrated signaling bias for calcium over cAMP compared to 32.
However, no signaling bias was observed in the cyclohexyl series with β-amino acid
1
6

derivative 25. Binding affinities of select compounds were also evaluated. Compounds 27 (K
i
=
0.187 µM), 28 (K
i
= 0.134 µM), 31 (K
i
= 0.571 µM), 34 (K
i
= 0.449 µM) competitively displaced
[125
I]-apelin-13. Compounds 27, 28, 31, 34 were also selective over the off-target AT1 receptor
(Table 3).
Table 4. Modification at N1 position of pyrazole core
APJ
APJ
APJ
APJ
AT1
Calcium
cAMP
β-arrestin
Binding
Calcium
Compd.
R
EC50
*
EC50
*
EC50
*
K
i
*
EC50*
µM±SEM
µM±SEM
µM±SEM
µM±SEM),
(µM)
(
%Emax
)
(%Emax
)
(%Emax
)
(%Emax
)
0
.344 ± 0.004 (91)
0.422 ± 0.008 (75)
ND*
ND
3
3
5.
6.
0.425 ± 0.06 (45)
0.049 ± 0.004 (91)
0.107 ± 0.046 (110)
0.282 ± 0.08 (97)
0.097 ± 0.032 (106)
0.078 ± 0.02 (77)
0.060± 0.03
0.054 ± 0.01
0.059 ± 0.02
>10
>10
>10
2
3
5.
7.
0.081 ± 0.025 (91)
0.070 ± 0.009 (94)
0.144 ± 0.008 (90)
0.063± 0.004 (89)
3
8.
0.101 ± 0.001 (87)
0.151 ± 0.04 (101)
0.092 ± 0.007 (78)
0.147 ± 0.06
>10
*
*
i
EC50 and K values are averages of multiple experiments performed in duplicate unless otherwise stated ± standard
error of the mean.
ND = Not Done
1
7

D. Modification at N1 position of pyrazole of the compound with cyclohexyl ethyl carboxyl amide
side chain:
As noted earlier in Table 1, modification at the N1 position of pyrazole core had been critical in
enhancing potency of carboxylic acid scaffold. To monitor the effects of similar N1 substituents on
cyclobutylamide scaffold, compounds 35 - 38 (Table 4) were synthesized. The cyclobutyl
carboxamide scaffold was chosen for modification since it showed better potency (Table 2) and was
considered more stable than the ester or other alkyl moieties. At first, substitution at N1 with aromatic
2
+
4
-fluorophenyl moiety 35 displayed moderate potency (Ca EC50 = 0.425 µM). However, alkyl and
cycloalkyl substituents 36, 25, 37, 38 significantly enhanced the potency of compounds compared to
2
+
their carboxylic acid derivatives 11-14. Specifically, compound 36 (Ca EC50 = 0.049 µM) and 37
2
+
(
Ca EC50 = 0.070 µM) with propyl and cyclopentyl substitution displayed 8-fold and 6-fold
2
+
enhanced potency compared to the corresponding carboxylic acid derivative 11 (Ca EC50 = 0.416
2
+
µM) and 13 (Ca EC50 = 0.424 µM). Thus, modification at N1 position resulted in potent compounds
5, 36, 37, 38 that exhibited EC50 values ≤ 100 nM. However, these compounds were not functionally
selective. Select potent compounds were also evaluated in radioligand binding assay.
Compounds 25 (K = 0.054 µM), 36 (K = 0.060 µM), 37 (K = 0.059 µM), 38 (K = 0.147 µM)
2
i
i
i
i
1
25
exhibited high binding affinities and competitively displaced [ I]-apelin-13.
Human liver microsomal stability studies
Our long-term goal is to develop potent, metabolically stable APJ biased agonists for chronic in vivo
studies to determine the importance of individual signaling cascades. Early in vitro assessment of
metabolic stability of compounds is necessary to determine possible in vivo metabolic liabilities.
Liver microsomal assays that includes cytochrome P450 system are the primary assays used to
evaluate the stability of compounds. A small set of potent compounds (Table 5) were therefore
evaluated for their stability in human liver microsomes. Our lead compound 9 demonstrated excellent
1
8

stability (t1/2 = 229 min, CLINT = 5 mL/min/kg). All the other analogs tested except 31 (t1/2 = 34.71
min, CLINT = 36 mL/min/kg) with 2-butanone substitution are having very poor stability as indicated
by high clearance values. This may be because all the other analogs contain alkyl/cycloalkyl
substitution at N1 that possibly undergo oxidative metabolism at α-carbon attached to the N1
nitrogen. Excellent stability of compound 9 is possibly due to the presence of stable aromatic 4-
fluorophenyl moiety at N1 pyrazole.
Table 5: Microsomal stability of synthesized compounds
Half-life,
Clearance,
CLINT
1
/2
Compd.
t
(min)
(mL/min/kg)
7
-ethoxy-4-
3
.84
325
methylcoumarin
9
.
229.59
3.21
3.10
3.55
3.26
8.18
34.71
3.59
2.90
6.25
6.28
5
2
2
2
2
2
3
3
3
3
3
1.
2.
3.
5.
8.
1.
4.
6.
7.
8.
389
402
351
383
153
36
347
431
199
199
1
9

Effects of Modifications on Ligand Bias
As noted earlier, small molecule apelin receptor agonists have been generated that are G
3
0
protein biased and result in increased cardiac contractility in vivo. We tested the effect of these
site modifications on bias relative to the reference endogenous agonist Pyr-Apelin-13 by
calculating bias factors35,36 between assays for G protein signaling (calcium and cAMP) and β-
arrestin recruitment (Table 5). In general, two patterns were observed (Figure 5): compounds that
displayed bias towards calcium signaling compared to cAMP inhibition and β-arrestin recruitment
(
Figure 5A) and compounds that did not display significant bias (< 10-fold) between pathways
Figure 5B). For those compounds that displayed significant bias between calcium signaling and
(
cAMP or β-arrestin, bias factors generally ranged between 2 and 3 on a log scale (corresponding
to 100- to 1000-fold bias towards calcium signaling).
Table 6: Bias factors for a series of APJ agonists (shown in dimensionless units on a logarithmic
scale ± standard error of the mean). Positive values denote bias towards the first pathway listed
and negative values denote bias towards the second pathway listed (pathway 1: pathway 2).
Compd.
Bias
Ca:cAMP
Bias
Ca:arr
Bias
cAMP:arr
Pyr-Apelin-13
0 ± 0.136
0 ± 0.194
0.89± 0.239
0.952± 0.189
1.18± 0.153
1.13± 0.17
0 ± 0.180
9
.
0.793± 0.172
0.743± 0.229
0.549± 0.133
0.947± 0.173
0.606± 0.163
0.895± 0.132
1.75± 0.169
2.45± 0.149
2.27± 0.217
0.097± 0.243
0.209± 0.23
0.634± 0.157
0.182± 0.178
0.454± 0.189
0.241± 0.167
-1.24± 0.199
-0.498± 0.231
-0.709± 0.344
1
1
1
1
1
1
2
2
0.
1.
2.
3.
4.
9.
7.
8.
1.06± 0.157
1.14± 0.164
0.509± 0.185
1.95± 0.215
1.56± 0.392
2
0

2
3
3
3
9.
0.
1.
4.
2.73± 0.39
2.75± 0.376
2.72± 0.161
2.45± 0.17
2.82± 0.805
2.46± 0.32
2.14± 0.219
1.8± 0.193
0.095± 0.882
-0.288± 0.479
-0.584± 0.243
-0.65± 0.186
Figure 5. Radar plots of bias factors for (A) compounds that display bias towards calcium
signaling and (B) compounds that have minimal (< 10-fold) bias towards any signaling pathways
compared to the reference agonist Pyr-Apelin-13. Bias factors are shown on a logarithmic scale.
Bias Ca:cAMP
Bias Ca:cAMP
Bias cAMP:βarr
Pyr-Apelin-13
Bias Ca:βarr
Bias Ca:βarr Bias cAMP:βarr
2
7
Pyr-Apelin-13
12
14
9
28
31
13
11
29
30
34
10
19
Conclusions
Systematic modifications at N1 position of the pyrazole core as well as the amino acid side chain
resulted in potent small molecule agonists exhibiting EC50 values of ≤ 100 nM. Functional
selectivity was a feature of several compounds with a bias towards calcium mobilization over β-
arrestin recruitment. Modification at N1 position of the pyrazole core resulted in compounds 11,
1
3, 14 (n-propyl, cyclopentyl, cyclohexyl substitutions) that exhibited enhanced potency and
functional selectivity towards calcium/cAMP over β-arrestin signaling. Conversion of the
carboxylic acid to alkyl and cycloalkyl amides were well tolerated and enhanced potency.
2
1

Compounds 21, 22, 25 with cyclohexyl ethyl side chain exhibited potency ≤ 100 nM in calcium
assay were equally potent towards other signaling cascades (cAMP and β-arrestin). Modification
at carboxylic acid portion of compounds having a phenylethyl side chain resulted in compounds
2
9, 30, 31 that displayed significant bias toward calcium signaling over β-arrestin cascade (bias
factors between 2 (100-fold) and 3 (1000-fold) on a log scale). Similarly, compounds 27-31, 34
exhibited significant bias toward calcium signaling over cAMP cascade. Among the set of biased
compounds, compound 31 displayed better metabolic stability (t1/2 = 34.71 min, CLINT = 36
mL/min/kg) and also did not activate the off-target AT1 receptor. Interestingly, elongation of the
carboxylic acid by one methylene (β-amino acid) in 34 demonstrated signaling bias for calcium
over cAMP compared to the one methylene short carboxylic acid derivative 32. However, no
signaling bias was observed in cyclohexyl series with β-amino acid 25. Modification at N1
position of the pyrazole resulted in compounds 25, 36, 37, 38 that exhibited high potency across
all the three signaling cascade. Specifically, compound 37 with N1 cyclopentyl substitution
exhibited the highest potency (cAMP EC50 = 0.097 µM) inhibiting forskolin-induced cAMP
accumulation. Compound 37 (cAMP EC50 = 0.097 µM) also exhibited ~10-fold enhanced
potency compared to the starting lead 9 (cAMP EC50 = 0.844 µM). Lipinski’s “Rule of 5” have
been useful guides for predicting oral absorption of compounds. The calculated properties of 37
(clogP 5.5, 2 donors, 5 acceptors, TPSA of 94.5 and a MW of 550.7) suggested that this compound
might potentially be orally bioavailable. In conclusion, we have identified potent small molecule
apelin receptor agonist 37 including several others that can be suitable as tool compounds for in
vivo studies of apelin receptor upon additional characterization.
2
2

Experimental Section: Reagents and starting materials were obtained from commercial suppliers
and were used without purification. Reactions were conducted under N atmosphere using oven-
dried glassware. All solvents and chemicals used were reagent grade. Anhydrous tetrahydrofuran
THF), dichloromethane (DCM), and N, N-dimethylformamide (DMF) were purchased from Fisher
2
(
Scientific and used as such. Flash column chromatography was carried out using a Teledyne ISCO
Combiflash Rf system and Redisep Rf gold pre-packed HP silica columns. Purity and
characterization of compounds were established by combination of HPLC, TLC, LC-MS, and NMR
1
analytical technique described below. H NMR spectra were recorded on a Bruker Avance DPX300
(300 MHz) spectrometer using CHCl
3
-d, MeOH-d4, DMSO-d with tetramethylsilane (TMS) (0.00
6
ppm) as the internal reference. Chemical shifts are reported in ppm relative to the solvent signal
and coupling constants (J) values are reported in Hertz (Hz). Thin-layer chromatography (TLC)
was performed on precoated silica gel GF Uniplates from Analtech and spots were visualized with
UV light or I detection or phosphomolybdic acid stain. Low-resolution mass spectra were obtained
2
using a Waters Alliance HT/Micromass ZQ system (ESI). Optical rotations were measured on an
Auto Pol III polarimeter at the sodium D line. Analytical and preparative HPLC was performed on
an automated Varian ProStar HPLC system equipped with a Xterra® C18 RP18 (4.6  100 mm i.
d., 3.5 m) column, with a flow rate of 0.1 mL/min.; max = 254 nm; mobile phase A: H
2
O and
mobile phase B: CH CN. Purity of the target compounds was determined to be ≥95% by HPLC.
3
Chemical names were generated using ChemDraw Ultra (CambridgeSoft, version 10.0).
Procedure: Scheme 1
Ethyl 4-(2,6-dimethoxyphenyl)-2,4-dioxobutanoate (40): To a solution of sodium ethoxide
(
21% in EtOH) (5.4 mL, 14.37 mmol) was added dropwise a mixture of diethyl oxalate (1.85 mL,
3.690 mmol) and 2,6-dimethoxy acetophenone (2.45 g, 13.69 mmol) in anhydrous ethanol (15
mL). The resultant mixture was stirred at room temperature for 30 minutes, upon which yellow
1
2
3

suspension formed. The reaction mixture was heated to reflux for 4 h. The reaction was cooled to
room temperature. Ethanol was evaporated in vacuo. The resultant residue was triturated with
diethyl ether (30 mL) and filtered to obtain sodium salt of ethyl 4-(2,6-dimethoxyphenyl)-2,4-
+
dioxobutanoate as yellow solid (4.0 g, 97 %). MS (ESI) m/z: Calcd. for C14
H
16
O
6
280.09 [M] ,
-
found 279.3 [M-H] .
Ethyl 5-(2,6-dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxylate (41a)
Step 1: Preparation of Isobutylhydrazine trifluoroacetate: Isobutyraldehyde (1g, 13.87 mmol)
and tert-butyl carbazate (1.8 g, 13.87 mmol) in methanol (25 mL) was stirred at room temperature
for 1 h. The solvent was evaporated and the resulting solid was dried under in vacuo to give white
solid of (E)-tert-butyl 2-(2-methylpropylidene)hydrazinecarboxylate in quantitative yield. Sodium
cyanoborohydride (1.2 g, 20.13 mmol) was added portionwise to a mixture of the (E)-tert-butyl 2-
(
2-methylpropylidene)hydrazinecarboxylate (2.5 g, 13.42 mmol) in 75 % of aqueous acetic acid
25 mL) at room temperature. The resultant solution was stirred for 3 h at room temperature. The
(
reaction mixture was neutralized with 1N NaOH, extracted with CH
2
Cl (3 x 25 mL), washed with
2
saturated NaHCO
3
, dried with Na
2
SO
4
, filtered, and evaporated to give tert-butyl 2-
1
isobutylhydrazinecarboxylate as oil (2.40 g, 95 %). H NMR (CDCl
3
, 300 MHz) δ 0.93 (d, J=6.78
Hz, 6 H), 1.46 (s, 9 H), 1.64 - 1.82 (m, 1 H), 2.43 (br. s., 1 H), 2.67 (d, J=6.78 Hz, 2 H). MS (ESI)
+
+
m/z: Calcd. for C
9
H
20
N
2
O 188.15 [M] , found 189.3 [M+H] .
2
Trifluoroacetic acid (12 mL) was added dropwise to a solution of the tert-butyl 2-
isobutylhydrazinecarboxylate (2.4 g, 12.75 mmol) in CH Cl (12 mL). The reaction mixture was
2
2
stirred at room temperature for 1.5 h. The solvent was evaporated to give the trifluoroacetate salt
1
of the title compound as colorless oil in quantitative yield. HNMR (CDCl
3
, 300 MHz) δ 1.04 (dd,
2
4

J=9.04, 6.78 Hz, 6 H), 2.04 - 2.25 (m, 1 H), 3.02 (dd, J=6.97, 3.96 Hz, 2 H). MS (ESI) m/z: Calcd.
+
+
for C
Step 2: Preparation of ethyl 5-(2,6-dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxylate
41a): Sodium salt of ethyl 4-(2,6-dimethoxyphenyl)-2,4-dioxobutanoate (40) (1.2 g, 3.96 mmol)
4
H
12
N
2
88.10 [M] , found 89.4 [M+H] .
(
and isobutylhydrazine trifluoroacetate (0.962 g, 4.76 mmol) was mixed with glacial acetic acid (25
mL) and conc. HCl (0.6 mL). The reaction mixture was heated to reflux for 3.5 h. After cooling,
reaction mixture was poured into water (25 mL). The aqueous layer was extracted with CH
x 30 mL) and the combined CH Cl layer was washed with saturated aqueous NaHCO
organic layer was then washed with saturated brine, dried over Na SO , followed by filtration. The
solvent was evaporated in vacuo. The residue was purified by silica gel flash chromatography (0-
0% EtOAc:Hex) to give major N -substituted title compound as colorless oil (535 mg, 40 %) and
less polar N -substituted, 2-isobutyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-carboxylate as
2
Cl
2
(3
2
2
3
. The
2
4
4
1
2
colorless oil (190 mg, 14 %).
Ethyl 5-(2,6-dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxylate (41a)
1H NMR (CDCl
, 300 MHz) δ 0.72 (d, J=6.78 Hz, 6 H), 1.39 (t, J=7.15 Hz, 3 H), 2.10-2.24 (m,
3
1
6
3
H), 3.72 (d, J=6.0 Hz, 2 H), 3.74 (s, 6 H), 4.41 (q, J=7.16 Hz, 2 H), 6.62 (d, J=8.67 Hz, 2 H),
+
.73 (s, 1 H), 7.38 (t, J=8.48 Hz, 1 H). MS (ESI) m/z: Calcd. for C18
H
24
N O
2 4
332.17 [M] , found
33.4 [M+H]+.
Ethyl 5-(2,6-dimethoxyphenyl)-2-isobutyl-1H-pyrazole-3-carboxylate (41b)
1H NMR (CDCl
3
, 300 MHz) δ 0.94 (d, J=6.78 Hz, 6 H), 1.37 (t, J=7.16 Hz, 3 H), 2.26 (m, 1 H),
3
.76 (s, 6 H), 4.33 (q, J=7.16 Hz, 2 H), 4.44 (d, J=7.16 Hz, 2 H), 6.61 (d, J=8.67 Hz, 2 H), 6.92
+
(s, 1 H), 7.28 (t, J=8.48 Hz, 1 H). MS (ESI) m/z: Calcd. for C18
H
24
N
2
O 332.17 [M] , found 333.3
4
[
M+H]+.
2
5

Ethyl 1-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-carboxylate (42a)
Step 1: Preparation of Cyclopentylhydrazine trifluoroacetate: Cyclopentanone (2.0 g, 23.77
mmol) and tert-butyl carbazate (3.1 g, 23.77 mmol) in methanol (20 mL) was stirred at room
temperature for 3 h. The solvent was evaporated and the resulting solid was dried under in vacuo
to give white solid of Tert-butyl 2-cyclopentylidenehydrazinecarboxylate (4.6 g, 98 %). Sodium
cyanoborohydride (2.1 g, 34.06 mmol) was added portionwise to a mixture of tert-butyl 2-
cyclopentylidenehydrazinecarboxylate (4.5 g, 22.697 mmol) in 75 % of aqueous acetic acid (25
mL) at room temperature. The resultant solution was stirred for 3 h at room temperature. The
reaction mixture was neutralized with 1N NaOH, extracted with CH
2
Cl (3 x 30 mL), washed with
2
saturated NaHCO
3
, dried with Na
2
SO
4
, filtered, and evaporated to give tert-butyl 2-
1
cyclopentylhydrazinecarboxylate as oil (4.4 g, 97 %). H NMR (CDCl
3
, 300 MHz) δ 1.40-1.55
(m, 4 H), 1.46 (s, 9 H), 1.63-1.75 (m, 4 H), 3.45-3.58 (m, 1 H). MS (ESI) m/z: Calcd. for
+
+
C
10
H
20
N
2
O
2
200.15 [M] , found 223.3 [M+Na] .
Trifluoroacetic acid (24 mL) was added dropwise to a solution of the tert-butyl 2-
cyclopentylhydrazinecarboxylate (4.4 g, 21.97 mmol) in CH Cl (24 mL). The reaction mixture
2
2
was stirred at room temperature for 2.5 h. The solvent was evaporated to give the trifluoroacetate
1
salt of the title compound as colorless oil in quantitative yield. HNMR (CDCl
3
, 300 MHz) δ 1.58
–
1.70 (m, 2 H), 1.72- 1.92 (m, 4 H), 1.95-2.09 (m, 2 H), 3.60-3.78 (m, 1 H). MS (ESI) m/z: Calcd.
+
+
for C
5
H
12
N 100.10 [M] , found 101.2 [M+H] .
2
Step 2: Preparation of ethyl 1-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-
carboxylate (42a): Sodium salt of ethyl 4-(2,6-dimethoxyphenyl)-2,4-dioxobutanoate (40) (1.2 g,
3
.96 mmol) and cyclopentylhydrazine trifluoroacetate (1.3 g, 5.95 mmol) was mixed with glacial
acetic acid (25 mL) and conc. HCl (0.6 mL). The reaction mixture was heated to reflux for 3 h.
2
6

After cooling, reaction mixture was poured into water (50 mL). The aqueous layer was extracted
with DCM (3 x 30 mL). The combined DCM layer was washed with saturated aqueous NaHCO
The organic layer was then washed with saturated brine, and then dried over Na SO , followed by
filteration. The solvent was evaporated in vacuo. The residue was purified by silica gel flash
chromatography (0-40% EtOAc:Hex) to give N -substituted title compound (1.0 g, 73 %, major
product) as light yellow solid and less polar N
dimethoxyphenyl)-1H-pyrazole-3-carboxylate as colorless oil (50 mg, 5 %).
3
.
2
4
1
2
-substituted,
2-cyclopentyl-5-(2,6-
1
1
-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-carboxylate (42a): H NMR (CDCl
3
,
3
00 MHz) δ 1.38 (t, J=7.2 Hz, 3 H), 1.44 - 1.57 (m, 2 H), 1.84 - 1.98 (m, 4 H), 2.08 - 2.26 (m, 2
H), 3.74 (s, 6 H), 4.21-4.31 (m, 1 H), 4.39 (q, J=6.9 Hz, 2 H), 6.63 (d, J=8.3 Hz, 2 H), 6.69 (s, 1
+
H), 7.38 (t, J=8.48 Hz, 1 H). MS (ESI) m/z: Calcd. for C19
H
24
N
2
O
4
344.41 [M] , found 345.0
[
M+H]+.
1
2
3
-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-carboxylate (42b): H NMR (CDCl
3
,
00 MHz) δ 1.36 (t, J=7.2 Hz, 3 H), 1.62 - 1.74 (m, 2 H), 1.85 - 1.98 (m, 2 H), 2.08 - 2.26 (m, 4
H) 3.77 (s, 6 H), 4.33 (q, J=7.16 Hz, 2 H), 5.68 (quin, J=7.49 Hz, 1 H), 6.62 (d, J=8.48 Hz, 2 H),
+
6
3
5
.91 (s, 1 H), 7.27 (t, J=8.48 Hz, 1 H). MS (ESI) m/z: Calcd. for C19
H
24
N O
2 4
344.41 [M] , found
45.3 [M+H]+.
-(2,6-Dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxylic acid (43): Lithium hydroxide
monohydrate (189 mg, 4.51 mmol) in 1 mL of water was added to a solution of ethyl 5-(2,6-
dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxylate (500 mg, 1.50 mmol) in MeOH (11 mL)
and THF (2 mL). The mixture was stirred at room temperature for 18 h. The reaction mixture was
concentrated to about half the volume and then extracted with ether (2 x 15 mL). The aqueous
layer was acidified with 1 N HCl and extracted with CH
2
Cl
2
(3 x 25 mL). The combined organic
SO . The solvent was evaporated in
layers were washed with water, brine and then dried with Na
2
4
2
7

1
vacuo to give the title compound as white solid (440 mg, 96 %). H NMR (CDCl
3
, 300 MHz) δ
0
.74 (d, J=6.40 Hz, 6 H), 2.10-2.24 (m, 1 H), 3.72 (d, J=7.54 Hz, 2 H), 3.75 (s, 6 H), 6.63 (d, J=9.0
+
Hz, 2 H), 6.79 (s, 1 H), 7.40 (t, J=8.48 Hz, 1 H). MS (ESI) m/z: Calcd. for C16
H
20
N
2
O
4
304.14 [M] ,
-
found 303.3 [M-H] .
1
-Cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-carboxylic acid (44): Following the
procedure described in 43, compound 44 was obtained from ethyl 1-cyclopentyl-5-(2,6-
1
dimethoxyphenyl)-1H-pyrazole-3-carboxylate. 98 % yield; white solid; H NMR (CDCl
3
, 300
MHz) δ 1.48 - 1.64 (m, 2 H), 1.84 - 2.02 (m, 4 H), 2.03 - 2.20 (m, 2 H), 3.75 (s, 6 H), 4.30 (quin,
J=7.4 Hz, 1 H), 6.64 (d, J=8.3 Hz, 2 H), 6.76 (s, 1 H), 7.40 (t, J=8.3 Hz, 1 H). MS (ESI) m/z:
+
+
-
Calcd. for C17
H
20
N
2
O
4
316.14 [M] , found 317.2 [M+H] , 315.4 [M-H] .
(S)-Tert-butyl 5-cyclohexyl-3-(5-(2,6-dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-
carboxamido)pentanoate (45): 5-(2,6-Dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxylic
acid (43) (40 mg, 0.13 mmol) was dissolved in THF (2 mL). To the solution was added BOP
(
58 mg, 0.13 mmol) and triethylamine (0.055 mL, 0.39 mmol). The resulting mixture was stirred
at room temperature for 15 minutes. (S)-Tert-butyl 3-amino-5-cyclohexylpentanoate (37 mg,
.14 mmol) was added and stirred at room temperature for 3 h. THF was evaporated in vacuo,
water was added to the residue and the aqueous layer was extracted with CH Cl (3 x 30 mL).
The combined organic layers were washed with water, brine and then dried with Na SO . The
0
2
2
2
4
solvent was evaporated in vacuo to give the crude residue. The residue was purified by silica gel
flash chromatography (0-50 % EtOAc:Hex) to give (S)-tert-butyl 5-cyclohexyl-3-(5-(2,6-
dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-carboxamido)pentanoate (45) as foam (65 mg,
1
9
1%). H NMR (CDCl
3
, 300 MHz) δ 0.75 (d, J=5.65 Hz, 6 H), 0.81-0.96 (m, 2 H), 1.10 – 1.36
(m, 6 H), 1.47 (s, 8 H), 1.61 - 1.73 (m, 8 H), 2.12 (dt, J=13.75, 7.06 Hz, 1 H), 2.54 (d, J=5.65
Hz, 2 H), 3.58 - 3.67 (m, 2 H), 3.72 (s, 3 H), 3.74 (s, 3 H), 4.31 - 4.42 (m, 1 H), 6.61 (d, J=8.67
2
8

Hz, 2 H), 6.69 (s, 1 H), 7.23 (s, 1 H), 7.37 (t, J=8.48 Hz, 1 H). MS (ESI) m/z: Calcd. for
+
+
C
31
H
47
N
3
O
5
541.35 [M] , found 542.6 [M+H] .
(S)-Tert-butyl 5-cyclohexyl-3-(1-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-
carboxamido)pentanoate (46): Following the procedure described in 45, compound 46 was
obtained using 1-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-carboxylic acid (44) and
1
(
S)-tert-butyl 3-amino-5-cyclohexylpentanoate. 87% yield; yellow solid; H NMR (CDCl
3
, 300
MHz) δ 0.81 - 0.96 (m, 2 H), 1.10 - 1.38 (m, 8 H), 1.48 (s, 9 H), 1.59 - 1.77 (m, 8 H), 1.81 - 1.96
(m, 4 H), 2.00 - 2.16 (m, 2 H), 2.55 (d, J=5.46 Hz, 2 H), 3.73 (s, 3 H), 3.74 (s, 3 H), 4.21 - 4.30
(m, 1 H), 4.30 - 4.43 (m, 1 H), 6.62 (d, J=8.48 Hz, 2 H), 6.67 (s, 1 H), 7.25 (br. s., 1 H), 7.37 (t,
553.35 [M] , found 555.0 [M+H]+
+
J=8.38 Hz, 1 H). MS (ESI) m/z: Calcd. for C32
H N
47 3
O
5
(S)-5-Cyclohexyl-3-(5-(2,6-dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-
carboxamido)pentanoic acid (12): Trifluoroacetic acid (0.6 mL) was added dropwise to a
solution of the (S)-tert-butyl 5-cyclohexyl-3-(5-(2,6-dimethoxyphenyl)-1-isobutyl-1H-pyrazole-3-
carboxamido)pentanoate (45) (60 mg, 0.11 mmol) in CH
2
Cl (1.5 mL). The reaction mixture was
2
stirred at room temperature for 2.5 h. The solvent was evaporated to give the trifluoroacetate salt
1
of the title compound. 54 % yield; white solid; H NMR (CDCl
3
, 300 MHz) δ 0.74 (d, J=6.78 Hz,
6
H), 0.81-0.98 (m, 3 H), 1.11-1.40 (m, 6 H), 1.60 - 1.78 (m, 8 H), 2.02-2.17 (m, 1 H), 3.66 (d,
J=7.54 Hz, 2 H), 3.74 (s, 6 H), 4.29-4.44 (m, 1 H), 6.62 (d, J=8.67 Hz, 2 H), 6.72 (s, 1 H), 7.38 (t,
+
J=8.48 Hz, 1 H), 7.37-7.51(m, 1 H). MS (ESI) m/z: Calcd. for C27
86.5 [M+H]+.
H
39
N
3
O
5
485.29 [M] , found
4
(S)-5-Cyclohexyl-3-(1-cyclopentyl-5-(2,6-dimethoxyphenyl)-1H-pyrazole-3-
carboxamido)pentanoic acid (13): Following the procedure described in 12, compound 13 was
2
9