Discovery and optimization of indole and 7-azaindoles as Rho kinase (ROCK) inhibitors (Part-II)

Article abstract of DOI:10.1016/j.bmcl.2011.09.084

Therapeutic interventions with Rho kinase (ROCK) inhibitors may effectively treat several disorders such as hypertension, stroke, cancer, and glaucoma. Herein we disclose the optimization and biological evaluation of potent novel ROCK inhibitors based on substituted indole and 7-azaindole core scaffolds. Substitutions on the indole C3 position and on the indole NH and/or amide NH positions all yielded potent and selective ROCK inhibitors (25, 42, and 50). Improvement of aqueous solubility and tailoring of in vitro and in vivo DMPK properties could be achieved through these substitutions.

Full text of DOI:10.1016/j.bmcl.2011.09.084

Bioorganic & Medicinal Chemistry Letters 21 (2011) 7113–7118
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and optimization of indole and 7-azaindoles as Rho kinase
(ROCK) inhibitors (Part-II)
E. Hampton Sessions , Sarwat Chowdhury , Yan Yin, Jennifer R. Pocas, Wayne Grant, Thomas Schröter,
⇑
⇑
Li Lin, Claudia Ruiz, Michael D. Cameron, Philip LoGrasso, Thomas D. Bannister , Yangbo Feng
Translational Research Institute and Department of Molecular Therapeutics, The Scripps Research Institute, Scripps Florida, 130 Scripps Way, #2A1, Jupiter, FL 33458, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 1 October 2011
Therapeutic interventions with Rho kinase (ROCK) inhibitors may effectively treat several disorders such
as hypertension, stroke, cancer, and glaucoma. Herein we disclose the optimization and biological eval-
uation of potent novel ROCK inhibitors based on substituted indole and 7-azaindole core scaffolds. Sub-
stitutions on the indole C3 position and on the indole NH and/or amide NH positions all yielded potent
and selective ROCK inhibitors (25, 42, and 50). Improvement of aqueous solubility and tailoring of in vitro
and in vivo DMPK properties could be achieved through these substitutions.
Keywords:
Rho kinase
ROCK
Indole
Ó 2011 Elsevier Ltd. All rights reserved.
Azaindole
Kinase inhibitor
Pyrazole
A downstream effector of the small GTPase Rho, Rho kinase
(ROCK) is a serine–threonine kinase of the AGC-family that regu-
lates vital cellular processes such as cell motility, size, division, dif-
ferentiation, and contractility.1–4 Two isoforms of ROCK have been
identified—ROCK-I and -II that are highly homologous in their ki-
nase domain (92% sequence identity).2 Since ROCK inhibition can
trigger cytoskeleton remodeling and reduce cellular contractility,
it is generally believed that inhibition of ROCK can be of therapeu-
tic value in diseases such as hypertension and glaucoma.5–9
In an earlier study (Part-I),10 we reported a new series of potent
ROCK inhibitors based on indole 2- and 3-carboxamides and their 7-
azaindole analogs (inhibitors 1–4, Fig. 1). While these unsubstituted
ROCK inhibitors possess high enzyme and cell activity and good PKA
selectivity, we believe that substitutions on the indole C3 position
(for the 2-carboxamide series), and on the indole NH and/or amide
NH positions will provide opportunities to perturb and fine tune the
pharmaceutical properties, such as the aqueous solubility and the
DMPK properties. Different applications could be then developed from
these tailored properties. Our working model for the binding of these
inhibitors with ROCK-II is shown in Figure 2 (also see the docking mod-
el in Part-I).10 We surmised that additional substitutions might be tol-
erable based on this binding motif. For example, R1 or R2 on inhibitors
1–4 might engage in binding interactions with previously unexplored
regions of the ATP-binding domain of ROCK-II, such as with Asp-176 or
with the solvent. Exploration in these regions (R1 and R2) was thus
undertaken by preparing analogs of inhibitors 1–4 using slight modi-
fications in the synthetic routes previously developed.10
The general synthesis of di-substituted indole 2-carboxamides 9
and 7-azaindole-2-carboxamides 10 commenced from the
corresponding carboxylic acids 5a and 5b10 as illustrated in Scheme
1. HATU mediated amide coupling with various amines 6 followed
by N-alkylation of the indole ring NH of 7a and 7b furnished the di-
substituted indole/azaindole framework. Finally a Suzuki hetero-
arylation reaction with 1H-4-pyrazoleboronic acid pinacol ester
(8) afforded the desired indoles 9 and 10.
For the synthesis of di-substituted indole 3-carboxamides 14
and its 7-azaindole analogs 15, the indole or 7-azaindole ring NH
(N1) of the corresponding trifluoromethyl ketones10 11a and 11b
were first alkylated with various alkyl halides 12. A Suzuki cou-
pling followed by hydrolysis under basic conditions gave the de-
sired carboxylic acids. In the final step, amide formation with
appropriate amines 6, using HATU as the coupling reagent, gave
the desired indole-3-carboxamide 14 and the corresponding 7-aza-
indole-3-carboxamide 15 (Scheme 2).
Taking advantage of the nucleophilicity of the C3 position of the
indole nucleus, a Mannich reaction was used to access 3-substi-
tuted 2-indolecarboxamides by treating compounds 16a and 16b
with formaldehyde and an appropriate amine in acetic acid. A
Suzuki heteroarylation with pyrazole boronic ester 8 then fur-
nished the target compound 17 (Scheme 3).
⇑
Corresponding authors. Tel.: +1 561 228 2201 (Y.F.); tel.: +1 561 228 2206
(T.D.B.).
E-mail addresses: tbannist@scripps.edu (T.D. Bannister), yfeng@scripps.edu
(Y. Feng).
The effect of substituents R1 and R2 on the indole-2-carboxam-
ides was first examined (Table 1).11 As demonstrated by compounds
These authors contributed equally to this work.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.084

7114
E. H. Sessions et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7113–7118
O
OMe
OMe
O
H
N
HN
OMe
OMe
N
N
H
N
H
N
N
H
HN
1
2
O
O
H
N
N
HN
N
N
H
N
N
H
N
HN
N
H
3
4
Figure 1. Lead indole and 7-azaindole ROCK inhibitors.
O
Lys-121
Y
O
CF3
CF3
R1-X
12
OMe
O
Cs2CO3, DMF, rt
P-loop
N
N
X
Y
N
H
X
Y
HN
R1
N
N
R2
11a Y=CH, X=Br
11b
13a Y=CH, X=Br
Met-172
R1
Y=N, X=Cl
13b
Y=N, X=Cl
Glu-170
kinase hinge
6
7
Y = CH, N
1
-
p
8
1) , PdCl2(PPh3)2, Na2CO3,
s
A
O
dioxane /H2O 110 °C (13a)
N
8
R2
OMe
Figure 2. The indole/azaindole ROCK inhibitor scaffold and proposed interactions
within the ROCK-II ATP-binding pocket.
or , Pd(PPh3)4, K2CO3,
EtOH/Tol 3:2, 140 °C (13b)
N
Y
N
HN
R1
2) 20% NaOH, reflux, 14 h
3) 6, HATU, Et3N, DMF, rt
14
Y=CH
18, 19, 21, and 22, the N1 substitution of the indole nucleus was not
tolerated by either ROCK isoforms with much reduced activity as
compared to compound 1. On the other hand, substitution on the
amide NH (compounds 20, 23, and 24) maintained a good ROCK
binding affinity. Nevertheless, R2 substitution did not lead to im-
proved selectivity for PKA nor did it improve ROCK potency.
15 Y=N
Scheme 2. A general route for the synthesis of 3-carboxamide indoles and
7-azaindoles.
contrast to the analogous indole 19, substitution by a larger group
with H-bonding capability still gave potent (slightly reduced ROCK
potency compared to 3 and 25, though) compounds (28 and 29).12
Compounds with substitutions at both R1 and R2 were found to
have low affinity for both ROCK isoforms (30 and 31), and the
one with a positively charged side chain (30) gave even lower
ROCK activity. These results indicate that perhaps some azaindole
2-carboxamides can bind in an orientation somewhat different
from the corresponding indole analogs.
A divergent trend of the substituent effect (relative to the indole
series, Table 1) was observed for the isosteric 7-azaindole-2-
carboxamide series. Similar to the indole 2-carboxamides, most
substitutions at the amide NH (R2) yielded potent ROCK inhibitors
with good PKA selectivity (32–34). The same was true when a
small methyl group was attached to the indole NH position (25).
However, introduction of large groups without H-bonding capabil-
ity, such as a methyl cyclopropyl or an aromatic group, at the in-
dole NH position was not tolerated (26 and 27). Remarkably, in
OMe
HN
R2
OMe
O
6
R2
N
N
H
OH
X
Y
HATU, Et3N,
DMF, rt
5a
5b
Y=CH, X=Br
Y=N, X=Cl
N
H
O
X
Y
7a
7b
Y=CH, X=Br
Y=N, X=Cl
1) R1-X, Cs2CO3, DMF
O
OMe
R2
N
HN
2)
B
8
N
O
PdCl2(PPh3)2, Na2CO3,
dioxane/water, 110 °C for 7a
or 8, Pd(PPh3)4, K2CO3,
N
O
Y
N
HN
R1
9
Y=CH
10
Y=N
EtOH/Tol 3:2, 140 °C for 7b
Scheme 1. A general route for the synthesis of 2-carboxamide indoles and 7-azaindoles.

E. H. Sessions et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7113–7118
7115
MeO
MeO
R4R3N
1) R3R4NH , CH2O, AcOH, 80 °C
HN
O
HN
O
2) 8, PdCl2(PPh3)2, Na2CO3
dioxane/ H2O 110 °C (16a)
or 8, Pd(PPh3)4, K2CO3,
EtOH/Tol3:2, 140 °C (16b)
N
H
N
H
Y
Y
X
HN
N
17a Y=CH
17b
16a
16b
Y=CH, X=Br
Y=N, X=Cl
Y=N
Scheme 3. Synthesis of 3-aminomethyl-2-carboxamide indoles and 7-azaindoles.
Table 1
Effect of N-substitutions on indole and 7-azaindole-2-carboxamides
R2
N
OMe
HN
N
X
N
O
R1
R1
H
R2
IC50 (nM)
a
Compd
X
ROCK-II
2
ROCK-I
19
PKA
1
CH
CH
H
H
6700
18
24
88
1800
OH
19
CH
H
310
710
>20,000
N
N
20
21
CH
CH
H
Me
Me
Me
3
200
37
1,100
2800
>20,000
22
23
CH
CH
Me
2300
11
6900
62
>20,000
1640
H
H
N
N
24
CH
10
46
1600
3
25
N
N
H
Me
H
H
2
<1
12
2
140
150
26
27
N
H
150
280
5000
OMe
N
H
4400
>20,000
Ndb
28
29
30
N
N
N
H
H
43
24
243
88
3670
5800
3700
N
OH
Me
1900
6600
N
31
32
33
N
N
N
Me
H
220
4
1000
27
2200
570
N
Me
H
8
33
1700
34
N
H
18
121
>20,000
N
a
Average of two or more measurements with derivations 630%.
Not determined.
b
We then turned our attention to the N-substitution effect on the
NH was always well tolerated and gave potent ROCK inhibitors
(35, 45–47). In contrast to the 2-carboxamide series, di-substitu-
tion at both the indole NH and the amide NH positions, with at
least one of the substitutions being a small group such as a methyl
group, still gave potent ROCK inhibitors (compounds 37, 39, 40,
and 43). However, di-substitution with two large groups yielded
poor ROCK binding (compound 44). Also noteworthy is that the
3-carboxamide inhibitors 2 and 4 (Table 2). The most important
observation is that many substitutions at the indole NH position
in both the indole and the 7-azaindole series were well tolerated.
These substitutions include small or large groups with or without
H-bonding capability (compounds 36, 38, 41, 42, 48, and 49).13 Like
that in the 2-carboxamide series, mono-substitution at the amide

7116
E. H. Sessions et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7113–7118
Table 2
use of polar H-bonding side chains at either (but not both) position
led to potent ROCK-II inhibitors with markedly reduced PKA activ-
ity (39–43, 48, and 49). For example, inhibitors 40 and 49, both of
which possess a tertiary amine side chain, were found to have
much reduced affinity for PKA compared to that of compounds 2
and 4, respectively.
Effect of N-substitutions on indole and 7-azaindole-3-carboxamides
O
R2
N
HN
N
The tolerance for large groups at the indole N1 position in 3-
carboxamides contrasts remarkably with the absence of clearly po-
sitive effects for such N1 substitutions in the 2-carboxamide series
(compareinhibitors 42, 43 with 19, 22, for example). This divergence
can be explained well by our docking studies (Fig. 2 in Part-I) that
suggest that the binding conformationof 2-carboxamide indoles dif-
fer significantly from their 3-carboxamide analogs with a flipped
orientation ofthe indole nucleus.10 In indole 2-carboxamides, the in-
dole NH is projected away from residue Asp-176 of ROCK-II and
toward the inside of the binding pocket. In such an orientation, there
is no space to accommodate a large alkyl group in that area. How-
ever, substitutions containing H-bonding capability at this position,
such as those in compounds 28 and 29 in Table 1, might pick up some
binding energy with the extra H-bonding interactions and thus still
render ROCK affinity although the molecule conformation might be
twisted. In the other orientation, predominant for most if not all 3-
carboxamides, the indole NH is instead projected towards Asp-176
and to the solvent, where there is an open space to hold many differ-
ent substitutions. Therefore, simple small or large groups (com-
pounds 37 and 38), or large groups containing H-bonding
capability (compounds 41–43, and 49) are all well tolerated at this
indole NH position.
This binding hypothesis for 2-carboxamide indole/azaindole
based ROCK inhibitors also suggests that adding a polar side chain
of an appropriate length at the indole C3, rather than at the indole
N1, would confer good ROCK potency. Additionally, such substitu-
tions would change the inhibitor’s physicochemical properties
and thus have the potential to fine-tune the pharmaceutical pro-
files. This proposal was tested by preparing indole 2-carboxamides
with polar groups added at the desired indole C3 position. As shown
in Table 3, all polar side chain containing compounds (50–55)
showed good ROCK-II potency. The morpholine analog 51 was
found to have similar ROCK-II affinity as inhibitor 1 with improved
selectivity over PKA (6000-fold). The N-methyl piperazine analog
(52) and the homo-piperazine analog (53) gave even better ROCK
potency although the PKA activity of 53 was increased as compared
to 1. Interestingly, these C3 substitutions slightly reduced both the
ROCK activity and the selectivity against PKA in the 7-azaindole ser-
ies although it still produced potent ROCK inhibitors (compounds
54 and 55 compared to compound 3).
Some of the potent compounds from each indole and azaindole
series were evaluated for their cell based activity. As shown in Ta-
ble 4, all those compounds active in enzyme assays with IC50 val-
ues of <20 nM generally gave cell potency with IC50 values of less
than ꢀ200 nM in our cell-based myosin light chain bis-phosphor-
ylation assays (ppMLC),14 indicating that these compounds all pos-
sess good cell permeability.
To evaluate the drugability of these indole and azaindole based
ROCK inhibitors, promising compounds from Tables 1–3 were also
subjected to pharmacokinetic (DMPK) assays. The most important
observation from the in vitro DMPK (Table 4) experiments was that
any substitutions to the indole-2-carboxamide scaffold would re-
duce the CYP-450 enzyme inhibitions (compounds 20, 23, 25, 50–
53 compared to 1) while the resulting compounds still remained
a good human liver microsomal (HLM) stability (compounds 20,
23, 25, 50, and 53). The exceptions were the morpholine analog
51 and N-methyl piperazine analog 52. Remarkably, the dimethyl-
amino analog 50 (also coded as SR7309) was very stable to human
liver microsomes and had a low CYP-450 inhibition profile, which
indicated that further investigations were warranted for this
X
N
OMe
R1
R1
R2
IC50 (nM)
a
Compd
X
ROCK-
II
ROCK- PKA
I
2
35
36
37
CH
CH
H
H
H
Me
H
<1
2
<1
<1
<1
19
<1
3
510
4300
87
CH Me
CH Me
Me
680
38
39
CH
H
5
1
21
13
Ndb
N
CH Me
2810
40
41
42
CH Me
CH
1
1
9
3
5
2200
900
N
H
H
OH
CH
<1
7000
N
N
43
44
CH
CH
Me
9
110
>20,000
>20,000
2800
8900
N
N
45
46
N
N
H
H
6
30
76
>20,000
5000
Et
H
19
47
48
N
N
H
62
<1
260
<1
3700
720
N
N
Me
49
4
N
N
H
H
8
2
89
11
>20,000
8000
N
H
a
Average of two or more measurements with derivations 630%.
Not determined.
b
inhibitor considering its high ROCK potency (ROCK-II IC50 = 9 nM)
and low PKA activity (PKA IC50 = 1500 nM). The most noticeable
observation for the 7-azaindole-2-carboxamide ROCK inhibitors
was that any substitutions on the C3 position, on the indole NH,
and/or on the amide NH decreased the microsomal stability (com-
pare compounds 28, 32–34, 54, and 55 to compound 3). On the
other hand, no significant changes in CYP-450 inhibitions were ob-
served for these substituted compounds with the exception of com-
pound 32 which showed slightly increased inhibitions for CYP
isoforms 1A2, 2C9, and 2D6. Nevertheless, none of these 7-azain-
dole-2-carboxamides gave significant CYP inhibitions.
The unsubstituted indole-2- and indole-3-carboxamide series
were notorious for their high CYP inhibitions as demonstrated by
data obtained for the lead compounds 1 and 2 (Table 4). Therefore,
any structural modifications that can reduce these inhibitions, as
seen in the 2-carboxamide series, will benefit the drugability of
the indole-carboxamide based ROCK inhibitors. However, the sub-
stitution effect on the CYP-450 inhibitions is a little different in the
indole-3-carboxamide series (35–42) from that in the indole-
2-carboxamide series (20, 23, and 25). A methyl substitution on
either or both the indole and the amide NH positions (35–38), or
a phenyl substitution on the indole NH (38) did not significantly
change this inhibition profile although the amide substitution

E. H. Sessions et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7113–7118
7117
Table 3
were on the indole nitrogen (41 and 42), CYP inhibitions were still
high. Interestingly, compounds 41 and 42 were also the exception
regarding effects of substitutions on the microsomal stability of
these indole-3-carboxamide inhibitors. The half-life of 41 and 42
in human microsomes (HLM) was almost unchanged while that
of all other compounds (35–40) was significantly reduced com-
pared to compound 2. In the 7-azaindole-3-carboxamide series,
no clear trend was observed regarding the effect of substitutions
on the microsomal stability and the inhibition for selected CYP iso-
forms (compounds 45, 48, and 49 compared to compound 4).
A few lead ROCK inhibitors were also selected for in vivo phar-
macokinetic studies (Table 5). The data for compound 3, which
have already been reported in Part-I (compound 21 in Part-I) are
also listed in Table 5 for comparison. Data for compound 25 indi-
cated that a simple methyl substitution to the indole nitrogen of
the 7-azaindole-2-carboxamide scaffold did not affect much of
the in vivo DMPK profile with only some reduction in AUC values.
Compound 25, which had an overall good pharmacokinetic profile,
could be a good candidate for systemic applications by oral dosing.
However, a cyclopropyl substitution on the amide nitrogen (33)
significantly worsened the properties with a remarkable increase
in clearance (Cl) and volume of distribution (Vss), a reduction in
half-life (t1/2), and AUC values of both iv and po dosing, and a sig-
nificant decrease in Cmax of oral dosing. Similar to their unsubsti-
tuted analogs, compounds based on 3-carboxamides (42 and 49,
for example) normally gave poor DMPK properties due to their
generally short half-lives in liver microsomes. These compounds
had high clearance (Cl) and volume of distribution (Vss), low AUC
and Cmax values, and minimum oral bioavailability. The C3 mor-
pholinomethyl substituted 7-zaindole-2-carboxamide 51, again
due to its low microsomal stability, also exhibited a pharmacoki-
netic profile with high clearance and very low oral absorbance. It
was anticipated that many of these compounds would be rapidly
cleared and would therefore be undesirable for systemic therapy.
Effect of C3-substitutions on indole and 7-azaindole-2 carboxamides
R1
H
N
OMe
HN
N
X
N
H
O
Compd
X
R1
H
IC50 (nM)
a
ROCK-II
ROCK-I
19
PKA
6700
1
CH
CH
2
9
50
N
N
34
40
6
1500
18,000
1300
51
52
CH
CH
3
O
N
N
<1
53
3
CH
N
<1
2
8
12
39
385
8000
911
N
NH
H
54
N
N
N
N
O
8
55
N
16
91
6100
a
Average of two or more measurements with derivations 630%.
(35) did reduce the 3A4 inhibition and the indole NH substitution
did lower the 1A2 inhibition (36–38). On the other hand, substitu-
tions with H-bonding capability (39–42) gave mixed results. When
these substitutions were on the amide nitrogen (39 and 40), CYP
inhibitions were greatly reduced, but when these substituents
Table 4
Cell activity and in vitro DMPK properties for selected compounds
a
Compd
ppMLC cell assays IC50 (nM)
t1/2 (min) HLM
CYP-450 inhibition at 10 lM (%)
1A2
2C9
2D6
3A4
1
20
23
25
3
28
32
33
34
2
35
36
37
38
39
40
41
42
4
45
48
49
50
51
52
53
54
55
28
70
80
46
49
68
53
17
24
27
21
25
13
9
42
À7
À2
1
87
27
65
76
38
46
83
52
26
97
94
93
93
96
28
62
87
91
86
91
77
51
27
86
75
53
À26
9
81
9
87
2
206
25
78
48
32
55
52
30
52
51
50
53
58
76
31
52
32
53
12
25
23
69
48
70
86
81
60
28
59
40
22
65
27
40
43
72
38
87
70
93
31
36
63
70
46
56
68
44
19
86
75
51
34
46
Ndb
Ndb
25
4
26
32
16
12
87
70
23
23
36
4
Ndb
Ndb
4
Ndb
4
17
4
Ndb
53
10
14
8
21
20
15
9
13
21
68
8
34
3
Ndb
23
71
À6
5
À5
À2
9
18
46
62
41
À15
7
Ndb
Ndb
Ndb
Ndb
Ndb
130
40
1
50
1
Ndb
Ndb
Ndb
4
a
Average of two or more measurements with derivations 630%.
Not determined.
b

7118
E. H. Sessions et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7113–7118
Table 5
In vivo pharmacokinetic profiles for selected inhibitors in ratsa
Compd
Cl (mL/min/kg)
Vss (L/kg)
t1/2 (h)
AUC iv (
l
M Ã h)
AUC po (
lM Ã h)
Cmax po (nM)
Oral F (%)
25
33
3
42
49
51
4
21
2
42
114
32
0.5
1.2
0.3
3.2
10.0
2.0
1.8
0.9
2.0
1.2
1.5
1.4
11.5
2.2
22.4
1.0
0.3
1.2
7.9
1.0
14.4
0.02
0.0
1190
255
1800
0
35
22
32
1
0
0
0
0
0.0
a
Data was generated from three determinations with derivations 630%. Dosed at 1.0 mg/kg (iv) or 2.0 mg/kg (po).
However, local applications could be practical, for example, topical
applications of some compounds could result in increased aqueous
humor outflow and decreased intraocular pressure8,9 as demon-
strated in several of our studies for compounds based on other
scaffolds developed in our laboratory.15,16
Acknowledgments
We thank Professors William Roush and Patrick Griffin for their
support, and Dr. Yen Ting Chen for helping the preparation of this
manuscript.
To demonstrate the improvement of aqueous solubility by these
substitutions with polar side chains (or substitutions with H-bond-
ing capability), two pairs of compounds (1 vs 50 and 2 vs 40) were
subjected to solubility testing at both pH 5.5 and 7.4. Compounds 1
References and notes
1. Leung, T.; Manser, E.; Tan, L.; Lim, L. J. Biol. Chem. 1995, 270, 29051.
2. Matsui, T.; Amano, M.; Yamamoto, T.; Chihara, K.; Nakafuku, M.; Ito, M.;
Nakano, T.; Okawa, K.; Iwamatsu, A.; Kaibuchi, K. EMBO J. 1996, 15, 2208.
3. Wettschureck, N.; Offeremanns, S. J. Mol. Med. 2002, 80, 629.
4. Riento, K.; Ridley, A. Nat. Rev. Mol. Cell Biol. 2003, 4, 446.
5. Nagaoka, T.; Morio, Y.; Casanova, N.; Bauer, N.; Gebb, S.; McMurtry, I.; Oka, M.
Am. J. Physiol. Lung Cell. Mol. Physiol. 2004, 287, 665.
6. Oka, M.; Fagan, K. A.; Jones, P. L.; McMurty, I. F. Br. J. Pharmacol. 2008, 155, 444.
7. Tanihara, H.; Inatani, M.; Honjo, M.; Tokushige, H.; Azuma, J.; Araie, M. Arch.
Ophthalmol. 2008, 126, 309.
and 2 had low aqueous solubility at both pH values (60.01
contrast, compounds 42 and 50 exhibited excellent aqueous solu-
bility at pH 5.5 (P100 M) mainly due to the tertiary amine moi-
eties on their substitutions. In addition, compounds 42 and 50 still
lM). In
l
had fair to good solubility even at pH 7.4 (77 and 1.1
respectively).
lM,
In conclusion, SAR studies demonstrated that incorporation of
side-chains to the C3 position or to indole and/or amide nitrogen
atoms of the indole carboxamide scaffolds still generated potent
ROCK inhibitors with good selectivity against PKA. These substit-
uents might affect the ROCK potency, PKA selectivity, and more
importantly, they could modify the inhibitor’s physicochemical
properties, and in vitro and in vivo DMPK profiles, which is
not easy to achieve in the original unsubstituted analogs. Several
of our compounds have suitable properties which warrant fur-
ther investigations to develop potential therapeutics, either for
local or for systemic applications. For example, compounds 42
(also coded as SR6781), 49 (also coded as SR9165), and 51 (also
coded as SR7280) might be good for nonsystemic local applica-
tions such as for the treatment of glaucoma. These compounds
have much better aqueous solubility compared to their unsubsti-
tuted analogs due to the polar side chains. Their poor pharmaco-
kinetic properties (high clearance and minimum oral absorbance)
make these compounds suitable for local applications in order to
diminish potential safety concerns. The overall best compound
for systemic oral applications is compound 25 (also coded as
SR7583). This compound has a low clearance, a low volume of
distribution, a fair half-life, high AUC values, and a good bio-
availability value in rat pharmacokinetic assays. In addition,
compound 25 might possess higher brain penetration compared
to its unsubstituted analog, which is very important for CNS
applications. The specific brain penetration data together with
in vivo studies for various applications of these compounds will
be reported in due course.
8. Nakajima, E.; Nakajima, T.; Minagawa, Y.; Shearer, T. R.; Azuma, M. J. Pharm. Sci.
2005, 94, 701.
9. Rao, P. V.; Deng, P. F.; Maddala, R.; Epstein, D. L.; Li, C. Y.; Shimokawa, H. Mol.
Vis. 2005, 11, 288.
10. Chowdhury, S.; Sessions, E. H.; Pocas, J. R.; Grant, W.; Schröter, T.; Lin, L.; Ruiz,
C.; Cameron, M. D.; Schürer, S.; LoGrasso, P.; Bannister, T. D.; Feng, Y.; Bioorg.
Med. Chem. Lett. in press. doi:10.1016/j.bmcl.2011.09.083.
11. ROCK and PKA assays were performed as described in: (a) Schröter, T.; Minond,
D.; Weiser, A.; Dao, C.; Habel, J.; Spicer, T.; Chase, P.; Baillargeon, P.; Scampavia,
L.; Schürer, S. C.; Chung, C.; Mader, C.; Southern, M.; Tsinoremas, N.; LoGrasso,
P.; Hodder, P. J. Biomol. Screen. 2008, 13, 17; (b) Chen, Y. T.; Vojkovsky, T.; Fang,
X.; Pocas, J. R.; Grant, W.; Handy, A. M. W.; Schröter, T.; LoGrasso, P.; Bannister,
T. D.; Feng, Y. Med. Chem. Commun. 2011, 2, 73.
12. In many instances compounds reported herein were also evaluated versus a
selected panel of other kinases (e.g., MRCK
activity <5 M was seen.
a, p38, Akt, and JNK3,). No off-target
l
13. The N-aryl indole B (e.g., compound 38 in Table 2) was synthesized by a
copper-mediated coupling reaction:
O
O
O
O
Ar-I, CuI
NH2
Br
N
H
Br
N
H
N
N
H
B
A
NH2
Ar
.
14. Schröter, T.; Griffin, E.; Weiser, A.; Feng, Y.; LoGrasso, P. Biochem. Biophys. Res.
Commun. 2008, 374, 356.
15. Yin, Y.; Cameron, M. D.; Lin, L.; Khan, S.; Schröter, T.; Grant, W.; Pocas, J.; Chen, Y.
T.; Schürer, S.; Pachori, A.; LoGrasso, P.; Feng, Y. ACS Med. Chem. Lett. 2010, 1, 175.
16. Fang, X.; Yin, Y.; Chen, Y. T.; Yao, L.; Wang, B.; Cameron, M. D.; Lin, L.; Khan, S.;
Ruiz, C.; Schröter, T.; Grant, W.; Weiser, A.; Pocas, J.; Pachori, A.; Schürer, S.; Lo
Grasso, P.; Feng, Y. J. Med. Chem. 2010, 53, 5727.