Discovery of benzimidazole derivatives as orally active renin inhibitors: Optimization of 3,5-disubstituted piperidine to improve pharmacokinetic profile

Article abstract of DOI:10.1016/j.bmc.2018.04.051

We previously identified 2-tert-butyl-4-[(3-methoxypropyl)amino]-N-(2-methylpropyl)-N-[(3S,5R)-5-(morpholin-4-ylcarbonyl)piperidin-3-yl]pyrimidine-5-carboxamide 3 as a potent renin inhibitor. Since 3 showed unacceptably low bioavailability (BA) in rats, structural modification, using SBDD and focused on physicochemical properties was conducted to improve its PK profile while maintaining renin inhibitory activity. Conversion of the amino group attached at the 4-position of pyrimidine to methylene group improved PK profile and decreased renin inhibitory activity. New central cores with carbon side chains were explored to improve potency. We had designed a series of 5-membered azoles and fused heterocycles that interacted with the lipophilic S3 pocket. In the course of modification, renin inhibitory activity was enhanced by the formation of an additional hydrogen bonding with the hydroxyl group of Thr77. Consequently, a series of novel benzimidazole derivatives were discovered as potent and orally bioavailable renin inhibitors. Among those, compound 13 exhibited more than five-fold of plasma renin inhibition than aliskiren in cynomolgus monkeys at dose ratio.

Full text of DOI:10.1016/j.bmc.2018.04.051

Accepted Manuscript
Discovery of Benzimidazole Derivatives as Orally Active Renin Inhibitors: Op-
timization of 3,5-Disubstituted Piperidine to Improve Pharmacokinetic Profile
Hidekazu Tokuhara, Yasuhiro Imaeda, Yoshiyuki Fukase, Koichi Iwanaga,
Naohiro Taya, Koji Watanabe, Ray Kanagawa, Keisuke Matsuda, Yumiko
Kajimoto, Keiji Kusumoto, Mitsuyo Kondo, Gyorgy Snell, Craig A. Behnke,
Takanobu Kuroita
PII:
S0968-0896(17)32084-9
https://doi.org/10.1016/j.bmc.2018.04.051
BMC 14334
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
11 January 2018
25 April 2018
26 April 2018
Please cite this article as: Tokuhara, H., Imaeda, Y., Fukase, Y., Iwanaga, K., Taya, N., Watanabe, K., Kanagawa,
R., Matsuda, K., Kajimoto, Y., Kusumoto, K., Kondo, M., Snell, G., Behnke, C.A., Kuroita, T., Discovery of
Benzimidazole Derivatives as Orally Active Renin Inhibitors: Optimization of 3,5-Disubstituted Piperidine to
Improve Pharmacokinetic Profile, Bioorganic & Medicinal Chemistry (2018), doi: https://doi.org/10.1016/j.bmc.
2
018.04.051
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. 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.

Discovery of Benzimidazole Derivatives as Orally Active Renin Inhibitors: Optimization
of 3,5-Disubstituted Piperidine to Improve Pharmacokinetic Profile
a,c
a
a
a
Hidekazu Tokuhara, Yasuhiro Imaeda, Yoshiyuki Fukase, Koichi Iwanaga, Naohiro
a,c
a
a,d
a,e
a,f
Taya, Koji Watanabe, Ray Kanagawa, Keisuke Matsuda, Yumiko Kajimoto, Keiji
a
a,c
b,g
b,h
Kusumoto, Mitsuyo Kondo, Gyorgy Snell, Craig A. Behnke, and Takanobu
*,a
Kuroita
a
Pharmaceutical Research Division, Takeda Pharmaceutical Company Limited, 26-1,
Muraoka-Higashi 2-chome, Fujisawa, Kanagawa 251-8555, Japan
b
Takeda California, Inc., 10410 Science Center Drive, San Diego, California 92121,
United States.
ABSTRACT We previously identified 2-tert-butyl-4-[(3-methoxypropyl)amino]-N-(2-
methylpropyl)-N-[(3S,5R)-5-(morpholin-4-ylcarbonyl)piperidin-3-yl]pyrimidine-5-
carboxamide 3 as a potent renin inhibitor. Since 3 showed unacceptably low
bioavailability (BA) in rats, structural modification, using SBDD and focused on
physicochemical properties was conducted to improve its PK profile while maintaining
renin inhibitory activity. Conversion of the amino group attached at the 4-position of
pyrimidine to methylene group improved PK profile and decreased renin inhibitory activity.
New central cores with carbon side chains were explored to improve potency. We had
designed a series of 5-membered azoles and fused heterocycles that interacted with the
1

lipophilic S3 pocket. In the course of modification, renin inhibitory activity was enhanced
by the formation of an additional hydrogen bonding with the hydroxyl group of Thr77.
Consequently, a series of novel benzimidazole derivatives were discovered as potent and
orally bioavailable renin inhibitors. Among those, compound 13 exhibited more than five-
fold of plasma renin inhibition than aliskiren in cynomolgus monkeys at dose ratio.
KEYWORDS: renin inhibitor, crystal structure, bioavailability, piperidine, topological
polar surface area
*Corresponding author. Phone: +81-466-32-1113.
e-mail: takanobu.kuroita@takeda.com.
Current affiliation.
2

c
Axcelead Drug Discovery Partners, Inc., 26-1, Muraoka-Higashi 2-Chome, Fujisawa,
Kanagawa 251-0012, Japan.
d
PRA Health Sciences KK, 3-8, Doshomachi 2-chome, Chuo-ku, Osaka, Japan, 541-0045.
e
SCOHIA PHARMA, Inc., 26-1, Muraoka-higashi 2-chome, Fujisawa, Kanagawa 251-
8555, Japan.
f
SPERA PHARMA,1nc., 17-85, Jusohonmachi 2-chome, Yodogawa-ku, Osaka532-0024,
Japan.
g
Sapphire Energy, Inc., 9363 Towne Centre Drive, San Diego, California 92121, United
States.
h
10996 Torreyana Rd. Suite 280 San Diego, CA 92121.
INTRODUCTION
The renin–angiotensin–aldosterone system (RAAS) is a well-studied pathway, which
1
regulates mainly blood pressure (BP) and renal function. The control of BP is essential for
the prevention of cardiovascular outcomes such as heart failure, stroke, atherosclerosis,
2
and kidney disease. Aspartic protease renin exclusively cleaves the protein
3
angiotensinogen to release the physiologically inert angiotensin I (Ang I). Angiotensin-
converting enzyme (ACE) then converts Ang I to angiotensin II (Ang II), which is a potent
vasoconstrictor agent that acts through binding to the Ang II type I receptor (AT1) and a
stimulant of mineralocorticoid (aldosterone) secretion from the adrenal cortex resulting in
4
sodium and water retention. The effectiveness of RAAS cascade blockade is proven by
5
extensive clinical studies using ACE inhibitors and AT1 receptor blockers. On the other
3

hand, these agents stimulate elevation in plasma renin concentration and plasma renin
6
activity due to a feedback mechanism. The inhibition of renin, which controls the first and
rate-limiting step of the RAAS, should result in complete blockade of the angiotensin
7
signal even in situations of elevated circulating or tissue active renin. Thus, renin has long
been recognized as a desirable target for antihypertensive drugs.
The development of clinically effective renin inhibitors was a major challenge for
medicinal chemistry because of the difficulties in identifying suitable agents with the
required profiles such as high affinity for renin’s active site and sufficient bioavailability
8
(
BA) to permit chronic oral administration. Effort to identify clinically efficacious direct
renin inhibitors (DRIs) were made by many pharmaceutical companies. First generation
9
DRIs were peptide mimetics like 1 (zankiren), which suffered from too low a BA to
justify clinical testing. Second generation DRI 2 (aliskiren) was approved in the US in
10
2
007, despite modest oral BA in humans (BA = 2.6% in humans). This was an important
milestone in the history of RAAS blockade development and the first new class of
11
antihypertensive drugs in more than a decade. Based on the positive results reported for
12
2
, the interest in DRIs with a desirable PK profile for the control of hypertension has
intensified and encouraged us to identify a “best-in-class” third-generation DRI.
Compound 2 characterized by its four exposed hydrogen bond donors (HBD) and high
number of rotatable bonds, suffered from poor intrinsic cell permeability, which could be
2
13,14
caused in part by high topological polar surface area (TPSA) value (146 Å ).
In the
15
preceding paper, the optimization of a highly attractive small molecule lead series has
been described, yielding compound 3 (Figure 1). Despite compound 3 has only two HBD
4

and good membrane permeability, PK study revealed that 3 exhibited a low BA value in
animals (BA = <1% in rats). We hypothesized that by adjusting physicochemical
properties, such as TPSA, with keeping rein inhibitory activity by SBDD approach, we
would be able to refine PK profile and identify a “best-in-class” renin inhibitor. Previous
study revealed that the amino group of the piperidine ring on compound 3 forms a tight
hydrogen bonding with the catalytic site (Asp32 and Asp215) of the enzyme (Figure 2). In
addition, the isobutyl group occupies the S1 site. This hydrophobic interaction is essential
sp
for strong potency. In contrast, the S1’, S3, and S3 sites (which are occupied by a
morpholine group, tert-butylpyrimidine group, and methoxybutyl group, respectively) are
targetable for modification.
SAR investigation of compound 3 showed that the removal of the hydrogen bonding
donor, replacement of the amino group on the pyrimidine in the 4-position improved PK
1
4
profile. In this report, we describe the investigation of the S3site binder with considering
physicochemical properties, which lead to the discovery of a series of novel benzimidazole
derivatives.
Figure 1. Structure of previously reported renin inhibitors and 3.
5

Ser76
Thr77
Gly217
Asp32
Asp215
Figure 2. The X-ray crystal structure (left) and binding mode (right) of 3 with renin.
CHEMISTRY
The synthesis of pyrimidine derivatives is outlined in Scheme 1. Methoxypropanol was
deprotonated by sodium hydride and reacted with the common intermediate
13
chloropyrimidine 23 . Upon quenching with water, carboxylic acid 24 was obtained. It
was then converted to 4 in two steps. The pentylamine side chain was introduced to afford
26. The obtained ester 26 was then hydrolyzed, reacted with morpholine, and deprotected
to give 5 in three steps with a good yield.
a
Scheme 1. Synthesis of 4 and 5.
6

a
Reagents and conditions: (a) 3-methoxypropan-1-ol, NaH, DMF, rt, then water, rt,
6%; (b) morpholine, WSC, HOBt, Et N, DCE, rt, 81%; (c) 1 M HCl in EtOAc, rt, 38%;
3
3
(d) pentan-1-amine, DIEA, DMF, 80 °C, 95%; (e) (1) 0.5 M NaOH, MeOH, rt, quant.; (2)
morpholine, WSC, HOBt, DCE, rt, 82%; (f) 1 M HCl in EtOAc, rt, 98%.
The asymmetric synthesis of key intermediate 35 was investigated as shown in Scheme 2.
15
Hydrolysis of the cis-dimethyl ester 28 and subsequent treatment with Ac O provided
2
acid anhydride 30. A catalytic enantioselective methanolysis reaction was performed in a
1
6
manner reported by Deng to afford 3S,5R-isomer in excellent enantiomeric excess
95%ee) when 0.2 equivalent of (DHQD) AQN was used. Despite the excellent specificity
(
2
of the asymmetric methanolysis reaction, use of the precious reagent (DHQD) AQN is
2
problematic for large-scale experiments. Hence alternative methods for this methanolysis
reaction were investigated. In Deng’s report, quinidine was also demonstrated to be a
useful catalyst in the same asymmetric reaction. Quinidine is derived from a natural source
(the bark of the cinchona tree) in large quantity; therefore, it is easily available for large
7

scale use. One equivalent of quinidine was applied to the large scale synthesis of mono-
ester 31. Further optical resolution by recrystallization of its (R)-(+)-1-phenylethylamine
salt provided the desired mono-ester 31 in 97%ee. The carboxyl group of 31 was
converted to a protected amine group by Curtius rearrangement. After the modification of
ester moiety of 32 to morpholine amide 33, the desired intermediates 34 and 35 were
obtained in two steps with good yield.
a
Scheme 2. Asymmetric synthesis of the key intermediate 35.
a
Reagents and conditions: (a) K CO , MeOH, water, reflux, 97%; (b) Ac O, reflux, 99%;
2
3
2
(
c) (DHQD) AQN, MeOH, Et O, THF, –40 °C, 57%; (d) (1) quinidine, MeOH, THF, –
2
2
4
0°C, 93%; (2) (R)-1-phenylethylamine, EtOH then KHSO aq, 69%; (e) Et N, DPPA,
4
3
toluene, 100 °C, BnOH, Et N, 80 °C, 73%; (f) (1) 1 M NaOHaq, MeOH, rt, 90%; (2)
3
morpholine, WSC, HOBt, Et N, DMF, rt, quant.; (g) (1) Pd(OH) /C, H , MeOH, rt; (2) 2-
3
2
2
methylpropanal, NaH(OAc) , MeOH, rt, 65–71%.
3
Scheme 3 illustrates the synthesis of pyrazole derivatives 7a,b. Alkylation of 36a and
36b with 4-methoxybutyl methanesulfonate in the presence of cesium carbonate occurred
8

at the 2-position of the pyrazole ring to give 37a and 37b. The ester moieties of 37a and
7b were hydrolyzed to afford carboxylic acids 38a and 38b. Coupling of these acids with
3
the amine 35 obtained above was performed using N,N,N',N'-
tetramethylchloroformamidinium hexafluorophosphate (TCFH) in the presence of N,N-
1
7
diisopropylethylamine (DIEA). Finally, 39a and 39b were deprotected to afford the
desired 7a and 7b.
a
Scheme 3. Synthesis of 7a,b.
a
Reagents and conditions: (a) 4-methoxybutyl methanesulfonate, Cs CO , DMA, 65 °C,
2
3
7
6
5–82%; (b) LiOH·H O, EtOH, water, 65 °C, 91–97%; (c) 35, TCFH, DIEA, DCE, rt,
6–98%; (d) 4 M HCl in EtOAc, EtOAc, rt, 95–97%.
2
The imidazole ring was constructed by the synthetic route described in Scheme 4.
Amidine 40 was reacted with glycine tert-butyl ester and a subsequent ring closing
9

1
8
reaction was accomplished with phosphoryl chloride to yield 41a. A methoxybutyl side
chain was introduced onto the imidazole ring of the obtained 41a and commercially
available 41b to give 42a and 42b. The aldehyde groups of those compounds were
oxidized to afford 43a and 43b. After amidation with 35, the chloro group was removed
by catalytic hydrogenation in the presence of potassium acetate. Removal of a Boc group
provided the imidazole derivatives 8a and 8b, respectively.
a
Scheme 4. Synthesis of 8a,b.
a
Reagents and conditions: (a) (1) tert-butyl glycinate hydrochloride, Et N, DMF, 80 °C,
3
5
5%; (2) TFA, DCE, rt, then POCl , toluene, DMF, 100 °C, 22%; (b) 4-methoxybutyl
3
methanesulfonate, Cs CO , DMA, 90 °C, 52–99%; (c) NaClO , NaH PO , t-BuOH, 2-
2
3
2
2
4
methyl-2-butene, rt, 99–100%; (d) 35, TCFH, DIEA, DCE, rt, 64–84%; (e) Pd(OH) /C,
2
H , KOAc, MeOH, rt, 38–48%; (f) 1 M HCl in EtOAc, rt, 76–99%.
2
1
0

The pyrrole derivative 9 was synthesized according to the method in Scheme 5.
Commercially available 46 was alkylated to give 47. After hydrolysis of ester 47, the
desired 9 was obtained by amidation with 35, followed by treatment with hydrochloric
acid for deprotection.
a
Scheme 5. Synthesis of 9.
a
Reagents and conditions: (a) 4-methoxybutyl methanesulfonate, Cs CO , DMF, 60 °C,
2
3
75%; (b) 2 M NaOH, MeOH, quant.; (c) 35, TCFH, DIEA, DCE, rt, 48%; (d) 4 M HCl in
EtOAc, EtOAc, rt, 2 h, 82%.
Scheme 6 summarizes the synthesis of imidazole 10a, pyrazole 10b, and 1,2,3-triazole
1
9
1
0c. -Keto ester 51, a common intermediate, was synthesized from 50. Rhodium-
catalyzed ring closure and then chlorination afforded imidazole 52a, which was converted
to 10a in the same manner giving rise to imidazole derivatives 8a and 8b. The pyrazole
1
1

intermediate 52b was obtained by treatment of 51 with N,N-dimethylformamide dimethyl
acetal (DMF-DMA), ring closure using phenylhydrazine, and hydrolysis of the ester
moiety. A 1,2,3-triazole ring was constructed by treating 51 with phenylazide and sodium
hydride. Then, the ester moiety was hydrolyzed without purification to afford 52c.
Condensation of these acids with amine 35 was performed by using TCFH. Deprotection
of the Boc group was carried out under acidic conditions to afford 10b and 10c.
a
Scheme 6. Synthesis of 10a-c.
a
Reagents and conditions: (a) (1) trimethyl orthoformate, H SO , MeOH, 65 °C, then 8
2
4
M NaOH, 95 °C, 81%; (2) oxalyl chloride, THF, 0 °C to rt, then Meldrum's acid, pyridine,
DCM, 0 °C, (3) MeOH, reflux, 89% for 2 steps; (b) (1) 4-(acetylamino)benzenesulfonyl
azide, Et N, MeCN, rt, 2 days, quant.; (2) 1-phenylurea, Rh (OAc) , toluene, DCE, 80 °C,
3
2
4
then TFA, rt, 92%; (3) POCl , 100 °C, 29%; (4) 1 M NaOH, MeOH, 80 °C, 87%; (c) (1)
3
DMF-DMA, toluene, 80 °C, then PhNHNH , EtOH, 80 °C, 77%; (2) NaOH, MeOH, rt,
2
1
2

quant.; (d) NaH, phenyl azide, DMF, rt, then 4 M NaOH, MeOH, 60 °C, 52%; (e) (1) 35,
TCFH, DIEA, DCE, rt, 2 h, 92%; (2) Pd(OH) /C, H , KOAc, MeOH, rt, 54%; (f) SOCl ,
2
2
2
DMF, toluene then 35, THF, DIEA, 28%; (g) 35, TCFH, DIEA, DCE, rt, 15 h, 75%; (h)
1–4 M HCl in EtOAc, rt, 91–99%.
In order to construct a 1,2,4-triazole ring, precursor 55 was synthesized in two steps
(
Scheme 7). 54 was acylated with ethyl chloroglyoxylate in the presence of triethylamine,
followed by treatment with Lawesson’s reagent gave the desired thioamide intermediate
5. Ring closure was achieved by heating with benzohydrazide in 1-butanol to afford 56 in
5
poor yield. The ester moiety was exchanged for a butyl group at the same time.
Conversion of 56 to 11 was accomplished in good yield.
a
Scheme 7. Synthesis of 11.
1
3

a
Reagents and conditions: (a) (1) Et N, ethyl chloroglyoxylate, THF, 0 °C to rt; (2)
3
Lawesson's reagent, toluene, 90 °C, 55% for 2 steps; (b) PhC(O)NHNH ,1-BuOH, 140 °C,
2
4
.4%; (c) (1) LiOH·H O, EtOH, water, rt; (2) 35, TCFH, DIEA, MeCN, rt, 74% for 2
2
steps; (d) 4 M HCl in EtOAc, rt, 75%.
Syntheses of imidazopyridine 12 and tetrahydroimidazopyridie 14 were depicted in
Scheme 8. Introduction of methoxybutyl group to imidazopyridine 58 was accomplished
by Wittig reaction. Palladium catalyzed hydrogenation of olefin 60 reduced both the olefin
moiety and the imidazopyridine core to give 61b. Selective olefin reduction was achieved
with palladium catalyst poisoned by diphenylsulfide to yield 61a. Esters 61a and 61b were
converted to the desired 12 and 14.
a
Scheme 8. Synthesis of 12 and 14.
1
4

a
t
Reagents and conditions: (a) KO Bu, THF, –78 °C to rt, 68%; (b) Pd/C, Ph S, H ,
2
2
EtOAc, rt, 49%; (c) Pd/C, H , EtOAc, rt, 80%; (d) (1) LiOH·H O, EtOH, 60 °C, then 34,
2
2
TCFH, DIEA, MeCN, rt, 62%; (2) 2 M NaOH, MeOH, 60 °C, quant.; (3) morpholine,
WSC, HOBt, DIEA, DMF, rt, 85%; (e) LiOH·H O, 8 M NaOH, EtOH, 60 °C, then 35,
2
TCFH, DIEA, MeCN, rt, 22%; (f) 4 M HCl in EtOAc, EtOAc rt, 89–97%.
Scheme 9 showed the syntheses of benzimidazole derivatives 13, 16–18 and 20–22.
Coupling reactions of 63 and 35 or 34 under basic conditions afforded 64a and 64b,
respectively. Alkyl groups were introduced into the NH group of the benzimidazole ring
using cesium carbonate to afford 65a–d. The ester moiety of 65c was converted to
morpholine amide group in two steps affording 66. Reduction of ester groups was
accomplished by calcium borohydride to give 67 and 69. The reaction of ester 65d with
methyl magnesium bromide yielded 68. Boc deprotection was carried out by using
1
5

hydrogen chloride in ethyl acetate. As for 67 and 69, methanol solution of hydrogen
chloride was used to prevent the esterification of the hydroxymethyl moiety.
a
Scheme 9. Synthesis of 13, 16–18 and 20–22.
a
Reagents and conditions: (a) 34 or 35, THF, water NaHCO , rt, 77–84%; (b) 4-
3
methoxybutyl methanesulfonate or MeI or EtI, CsCO , DMF or DMA, 50–70 °C, 80–
3
8
7%; (c) (1) 1 M NaOH, MeOH, THF, rt, quant.; (2) morpholine, WSC, HOBt, Et N,
3
MeCN, rt, 63%; (d) NaBH , CaCl , EtOH, THF, 0 °C to rt, 47–63%; (e) MeMgBr, THF,
4
2
–40 °C to rt, 54%, (f) 4 M HCl in EtOAc, rt, 76–99%; (g) 10–20% HCl in MeOH, rt, 91–
99%.
Scheme 10 described the synthesis of indole 15. Displacement of indole 70 with 4-
methoxybutyl methanesulfonate and cesium carbonate, followed by hydrolysis afforded 71.
Subsequent amide coupling and deprotection of 71 provided 15.
a
Scheme 10. Synthesis of 15.
1
6

a
Reagents and conditions: (a) (1) 4-methoxybutyl methanesulfonate, Cs CO , DMF, 60
2
3
°C; (2) 4M NaOHaq 80 °C, 82% for 2 steps; (b) (1) 35, TCFH, DIEA, DCE, rt, 19%; (2)
4
M HCl in EtOAc, EtOAc, rt, 89%.
Synthesis of 19 was shown in Scheme 11. Mono-alkylation of phenylenediamine 72 gave
3. The benzimidazole ring was constructed by the reaction of 73 with methyl 2,2,2-
7
trichloroethanimidate to provide 74. The desired 19 was obtained by the coupling reaction
20
of 74 and 75 , followed by the deprotection.
a
Scheme 11. Synthesis of 19.
1
7

a
Reagents and conditions: (a) 4-methoxybutyl methanesulfonate, K CO , MeCN, reflux,
2
3
5
6%; (b) CCl3 C(NH)Ome, AcOH, rt, 96%; (c) 75, MeCN, water K CO , 80 °C, 63%;
2
3
(d) 4 M HCl in EtOAc, EtOAc, rt, quant..
RESULTS AND DISCUSSION
The compounds listed in Tables 1–5 inhibited recombinant human renin activity in the
absence or presence of human serum albumin (HSA– or HSA+, respectively) and
endogenous renin activity in human plasma (human plasma renin activity, hPRA). These
inhibitory activities were evaluated and reported as IC values.
50
As the first step toward our goal, the effect for the potency and PK profiles by replacing
the amino group on the pyrimidine ring at the 4-position was investigated to know the
contribution of HBD (Table 1). Although the renin inhibitory activity of oxa-analogue 4
13
was diminished, the carba-analogue 6 exhibited moderate renin inhibitory activity. These
decreases in renin inhibitory activity could be rationalized by the loss of hydrogen bonding
to the Gly217 backbone oxygen of the renin protein observed in X-ray co-crystal studies.
In particular, repulsion between the lone pair of the ether oxygen atom of 4 and the
1
8

carbonyl group of Gly217 caused the disappearance of renin inhibitory activity. The
importance of the oxygen atom in the methoxy group was also investigated. Although,
under HSA+ conditions, the renin inhibitory activities of 3 and 5 were similar, 5 showed
weaker hPRA inhibition than 3 (5.4 nM and 0.79 nM, respectively). The lipophilicity of
compounds (LogD: 4.45 for 5 and 2.99 for 3) dramatically affected their activity in the
presence of proteins such as HSA and in plasma, which has relevance to in vivo potency.
Reducing the LogD value would thus be important for optimizing compounds to improve
their efficacy in vivo.
In the rat cassette dosing study, the high clearance value (7261 mL/h/kg) and undetected
AUC value of 3 indicated that the poor PK profile of 3 was due to rapid excretion and/or
po
poor absorption rather than to metabolic stability, which was shown to be good (data not
shown). On the other hand, 6 showed improved clearance value (2407 mL/h/kg) and
AUC value (20.9 ngh/mL), giving the improved BA (5.0%). Both 3 and 6 possessed
po
good parallel artificial membrane permeability assay (PAMPA) permeability. The
improved PK profile of 6 seemed to be due to a difference in the rate of drug excretion
from the bloodstream. The reduced TPSA value of 6 would explain the difference of
2
2
clearance of compound (TPSA: 87.7 Å for 6 and 109 Å for 3). Lowering the TPSA of
compound was seemed the key factor for the orally bioavailable renin inhibitors. These
results inspired us to prepare a new S3 site binder with methoxybutyl side chain for the
recovery of potency.
Table 1. SAR and ADME profiles of pyrimidine derivatives 3–6.
1
9

a
b
IC (nM)
rat cassette dosing
PAMPA
pH7.4
5
0
TPSA
c
cmpd R
LogD AUC AUC CL
F
iv
ngh/mL)
d
po
2
HSA– HSA+ hPRA
(Å )
(
(mL/h/kg) (%) (nm/sec)
d
d
d
d
4
5
> 100 > 100 NT
2.81 NT
NT
NT
NT 272
106
0
.20
0.64
5.4
e
4.45 27.3
ND
3745
2407
7261
< 1 252
5.0 435
< 1 390
99.7
(0.15– (0.36– (3.5–
0
.26) 1.1)
8.3)
> 10
0.79
2
9
24
6
4.77 42.0
2.99 14.1
20.9
87.7
109
(22–
(19–
31)
3
8)
1
.5
0.38
e
3
ND
(0.60– (0.24– (0.61–
3
.6) 0.62) 1.0)
a
IC values and 95% confidence limits are calculated from the concentration-response
50
b
curves generated by GraphPad Prism. Dose: 1 mg/kg, po, 0.1 mg/kg, iv. Average of 3
rats. F is the measure of bioavailability. Not tested. Not detected.
c
d
e
The vast lipophilic S3 site would accept various substituents and permitted us to
introduce different scaffolds. To explore more preferable S3 site binders, we initially
examined replacement of the 6-membered pyrimidine ring by a 5-membered azole (Table
2). Pyrazole derivatives 7a,b and imidazole derivatives 8a,b with tert-butyl or phenyl
group were initially synthesized as new S3 site binders. Compound 7a showed potent
2
0

renin inhibitory activity under HSA– and HSA+ conditions, while potency of hPRA
inhibition was somewhat decreased, maybe due to the lipophilicity (LogD = 2.73).
Interestingly, changing tert-butyl to a phenyl group resulted in the disappearance of
potency (7a vs 7b). On the other hand, the imidazole derivative 8a showed moderate
potency under each of the conditions and 8b possessed comparable potency to 7a,
indicating that the substitution manner of 8b can be used for further substituent
modification. The scaffold of 8b, on which a phenyl group is located vicinal to the
methoxybutyl side chain, was employed to further explore 5-membered azole cores.
Table 2. SAR of 5-membered azoles 7a,b and 8a,b.
a
IC (nM)
5
0
cmpd Ar
LogD
HSA–
.0
1.6–
HSA+
3.6
hPRA
2.8%
2
5
7
7
8
a
b
a
inhibition 2.73
at 10 nM
(
(2.5–
5.2)
2
.4)
b
>100
>100
NT
2.68
2.12
4
8
35
b
NT
(34–69) (25–50)
2
1

4
.4
3.3
2.8
8b
2.03
(
6
3.0–
.3)
(2.6–
4.3)
(1.2–7.0)
a
IC values and 95% confidence limits are calculated from the concentration-response
50
b
curves generated by GraphPad Prism. Not tested.
All 5-membered azole derivatives (9, 10a–c, and 11) listed in Table 3 showed potent
renin inhibitory activity in the absence of HSA. The level of hPRA inhibition of pyrrole 9
was decreased despite strong activity under HSA– conditions as reflected by higher LogD
value (3.08). The potency of imidazole 10a was increased as compared to that of 8b,
while the difference between these compounds is in the position of the nitrogen atom and
not in LogD value. Notably, the presence of the nitrogen atom next to the carbonyl group
resulted in relatively stronger hPRA inhibition (10b vs 10c and 8b vs 11, respectively).
The X-ray co-crystal structures of 9 and 10c revealed that the distance between the
hydroxyl group of Thr77 and 9 were separated by 3.7 Å (Figure 3). On the other hand, a
hydrogen bond interaction was observed between Thr77 and the nitrogen atom of 10c with
a distance of 2.6 Å. The rotation of Thr77 residue was observed and indicated that
formation of a hydrogen bonding between the nitrogen atom of 10c and Thr77 could
account for the increase in potency.
Rat cassette dosing study of synthesized compounds revealed that the oral absorption of
compounds 9 and 10c (AUC : 18.5 and 3.5 ngh/mL) was better than that of other
po
compounds, which is probably due to their better PAMPA membrane permeabilities (217
and 369 nm/sec). The compounds with low logD value tend to exhibit low membrane
2
2

2
permeability (10a, 10b, and 11). Compound 9, with the lowest TPSA (76.0 Å ), exhibited
lowest clearance value among the compounds in Table 3. Compound 10c had good in vitro
profiles, but its high clearance value resulted in an unacceptable PK profile for oral
2
administration. The high TPSA value (102 Å ) could be the reason for that unfavorable
drug excretion as in the case of compound 3. These results encouraged us to explore other
S3 site binders in more detail.
Table 3. SAR and ADME profiles of phenylazoles 9, 10a–c, 11.
a
b
IC (nM)
rat cassette dosing
PAMPA
pH7.4
5
0
T
c
Cmpd Ar
LogD AUC AUC CL
F
iv
po
HSA– HSA+ hPRA
(Å
(nm/sec)
(ngh/mL)
(mL/h/kg) (%)
0
.16
1.4
3.4
9
3.08 34.6
1.84 13.3
1.78 20.8
18.5
2968
7604
4826
5.4 217
< 1 78
0.3 78
76
88
88
(
0
0.082– (0.97– (1.6–
.3)
2.1)
1.1
6.8)
1.2
0
.83
d
1
1
0a
0b
ND
(0.59– (0.66– (0.93–
1
.2)
1.8)
1.5)
1
.1
1.2
1.3
0.7
(0.93– (0.93– (0.84–
1
.4) 1.6) 2.2)
2
3

1
.0
0.65
0.40
1
1
8
0c
1
1.94 18.4
1.79 15.3
2.03 17.6
3.5
ND
ND
5544
6740
5804
1.9 369
< 1 116
< 1 72
10
10
88
(
1
0.61– (0.50– (0.0044–
.7)
.4
0.86) 1.2)
1.8 1.2
1
d
d
(0.88– (1.2– (0.74–
2
.1)
2.6)
2.0)
4
.4
3.3
2.8
b
(3.0–
(2.6– (1.2–
4.3) 7.0)
6
.3)
a
IC values and 95% confidence limits are calculated from the concentration-response
50
b
curves generated by GraphPad Prism. Dose: 1 mg/kg, po, 0.1 mg/kg, iv. Average of 3
rats. F is the measure of bioavailability. Not detected.
c
d
Figure 3. Overlay of crystal structures of 9 (white) and 10c (green) bound to renin. The
solid yellow line indicates the hydrogen bonding in the structure of 10c. The dashed-
yellow line indicates the distance between hydroxyl group of Thr77 and carbon atom
measured within the structure of 9.
2
4

The strategy to further improve PK profile is adjusting physicochemical properties, such
as reducing the molecular weight (all the compound listed in Table 1–3 were greater than
500), TPSA and the number of rotatable bonds without losing the interaction as described
1
3,14
above.
Imidazopyridine 12 and benzimidazole 13 were designed on the basis of
concepts as follows: 1) the nitrogen atom of the new central core would interact with the
Thr77, 2) fused heterocycles would reside in the same lipophilic S3 site as 3 occupies, 3)
the molecular weight and number of rotatable bonds were reduced by forming a fused ring
as a new scaffold (Table 4). Both 12 and 13 exhibited potent hPRA inhibition and lowered
TPSA value. Although the imidazopyridine 12 showed moderate membrane permeability
(63 nm/sec), which is probably due to low LogD value (1.54), its oral absorption was
confirmed (AUC : 9.5 ngh/mL) and clearance from the bloodstream was mitigated (CL:
po
2
2
304 mL/h/kg), reflecting its TPSA value (88.4 Å ). Furthermore, 13 exhibited a
significant improvement in rat AUC (92.2 ngh/mL) and BA (14%) with good membrane
po
permeability (345 nm/sec). These results indicate our optimization strategy worked well.
Adjusting the lipophilicity and TPSA of compounds is \essential for good BA in this
chemotype. To improve poor absorption, imidazopyridine 12 was reduced to provide
compound 14. However, 14 showed three-fold weaker hPRA inhibition than 12. To
confirm the importance of the hydrogen bonding with Thr77, 15 was synthesized.
Interestingly, 15 exhibited strong potency under HSA– condition, but considerable
decrease of hPRA inhibition was observed. We concluded that this lost in potency could
be derived from the loss of the hydrogen bonding. Therefore, forming a suitable interaction
with renin protein is essential to overcome this gap in this scaffold.
2
5

The X-ray crystal structure of 13 bound to renin protein is shown in Figure 4. As
expected, the binding mode of 13 was similar to that of 3. Hydrogen bonding between the
nitrogen atom of the benzimidazole and Thr77 was observed.
Table 4. SAR and ADME profiles of fused heterocycles 12–15.
a
b
IC (nM)
rat cassette dosing
PAMPA
pH7.4
5
0
TPSA
c
cmpd Ar
LogD AUC AUC CL
F
iv
po
2
HSA– HSA+ hPRA
(Å )
(ngh/mL)
(mL/h/kg) (%) (nm/sec)
1.6
2.1
1.1
1
1
1
1
2
3
4
5
1.54 43.5
2.11 68.0
9.5
2340
1554
2.2 63
88.4
88.9
88.9
76
(
3
0.84– (1.2– (0.97–
.1)
3.5)
1.3)
1.4
0
.43
0.84
92.2
14 345
(0.29– (0.68– (1.1–
0
.66) 1.0)
2.7)
2
.2 2.7
3.3
d
d
d
d
d
d
NT
NT
NT
NT
NT NT
(1.0– (1.4– (1.5–
4
.4)
5.1)
8.0)
1
.5
0.81
d
d
> 10
2.66 127.9 NT
799
NT 282
(1.1– (0.64–
2.2) 1.0)
a
IC values and 95% confidence limits are calculated from the concentration-response
50
b
curves generated by GraphPad Prism. Dose: 1 mg/kg, po, 0.1 mg/kg, iv. Average of 3
rats. F is the measure of bioavailability. Not tested.
c
d
2
6

Thr77
Figure 4. Crystal structure of 13 bound to renin. The solid yellow line indicates the
hydrogen bonding in the structure of 13.
sp
Finally, reinvestigation of the S1’ and S3 site binders was conducted with the
benzimidazole core (Table 5). 16 maintained potency under both HSA– and HSA+
sp
conditions in spite of the absence of the S3 site binder, but hPRA inhibition of 16 was
decreased (IC = 4.0 nM). Introduction of methyl group increased hPRA inhibition (17,
50
IC = 2.9 nM), and ethyl analog showed comparable hPRA inhibition to 13 (18, IC =
50
50
sp
1
.3 nM). This finding showed that occupation of the S3 site was not necessary for the
potent hPRA inhibition in this benzimidazole scaffold. Rat cassette dosing study revealed
that compounds with small substituents, such as 16–18 did not show superior PK profiles
over 13, nonetheless, some acceptable PK profiles were observed. Next, we explored S1’
sp
site binders. As in the case of the S3 site binder, compound 19, which did not have
S1’site binder, showed comparable potency under HSA– and HSA+ conditions and
decreased hPRA inhibition. X-ray crystal structure study of 3 revealed that the carbonyl
2
7

group of the morpholine amide moiety forms hydrogen bonding interaction with Ser76.
Also, the morpholine moiety was occupying the S1’ site (Figure 2). Instead of
morpholinocarbonyl group, a hydroxyl group was introduced as a small substituent for the
better PK profiles. 20 exhibited comparable potency to 13, suggesting that a hydrogen
bonding interaction between the hydroxyl group of 20 and Ser76 was formed successfully.
However, 20 showed lower Log D and PAMPA value than 13; rat PK profile of 20 was
comparable to 13. To improve permeability by masking the polarity of hydroxyl group, the
conversion to the tertiary alcohol was attempted (21). The PAMPA value was improved as
expected, but hPRA inhibition was decreased. This suggested that the hydrogen bonding
interaction with Ser76 might be disrupted by the dimethyl group. Finally, combination
sp
study of the S3 and S1’ site binders was investigated by employing the hydroxymethyl
sp
group of 20 as the S1’ site binder, with the ethyl group as the S3 site binders. Contrary to
our expectation, 22 showed decreased renin inhibitory activity. This result indicated that
both of the hydrogen bonding interaction with Ser76 and occupation of the morpholine and
sp
methoxybutyl groups at the S1’ site and S3 sites had contributed to the strong potency
(18 and 20 vs 22). Utilizing both binding sites would be better to show potent renin
inhibition in human. All the benzimidazole compounds showed improved clearance value.
This fact supports that the strategy of lowering TPSA is effective to enhance BA by
improving the drug clearance from bloodstream. Compound 13 exhibited strong potency,
good membrane permeability and reliable PK profiles, and was selected for further in vivo
evaluation.
2
8

Table 5. SAR and ADME profiles of benzimidazole derivatives 16–22.
PAMP
A
a
b
IC (nM)
rat cassette dosing
50
TPS
A
cmp
d
Log AUC
1
2
c
R
R
AUCpo CL
F
pH7.4
D
iv
HSA HSA hPR
2
(
Å )
–
+
A
(
mL/h/k (% (nm/se
(ngh/mL)
g)
)
c)
0
.59
0
.76
4.0
(
–
0
0.42
1
1
6
7
H
1.74 50.7
41.7 2053
8.2 60
90.6
79.7
(
–
0.46
1.2)
(2.7–
6.0)
.85)
0
.63 0.66
2
.9
(
–
0
0.43 (0.46
Me
Et
1.71 43.9
74.2 2332
17 274
(2.0–
–
4
.4)
.93) 0.95)
0
.66 0.89 1.3
1
1
8
9
1.79 43.4
2.00 52.1
58.9 2312
59.7 1978
14 242
12 324
79.7
59.4
(
–
0.41 (0.64 (0.90
1.0) –1.2) –2.3)
0
.9 1.6 9.1
H
(
0.64 (1.3– (4.8–
–1.3) 1.9) 29)
0
.28 0.32
1
.5
(
–
0
0.22 (0.22
20
1.77 107.7 103.4 931
9.6 82
79.6
(
–
0.82
2.9)
–
.35) 0.47)
2
9