Synthesis and Evaluation of 4-Hydroxycoumarin Imines as Inhibitors of Class II Myosins

Article abstract of DOI:10.1021/acs.jmedchem.0c01062

Inhibitors of muscle myosin ATPases are needed to treat conditions that could be improved by promoting muscle relaxation. The lead compound for this study ((3-(N-butylethanimidoyl)ethyl)-4-hydroxy-2H-chromen-2-one; BHC) was previously discovered to inhibit skeletal myosin II. BHC and 34 analogues were synthesized to explore structure-activity relationships. The properties of analogues, including solubility, stability, and toxicity, suggest that the BHC scaffold may be useful for developing therapeutics. Inhibition of actin-activated ATPase activity of fast skeletal and cardiac muscle myosin II, inhibition of skeletal muscle contractility ex vivo, and slowing of in vitro actin-sliding velocity were measured. Several analogues with aromatic side arms showed improved potency (half-maximal inhibitory concentration (IC50) <1 μM) and selectivity (≥12-fold) for skeletal myosin versus cardiac myosin compared to BHC. Several analogues blocked neurotransmission, suggesting that they are selective for nonmuscle myosin II over skeletal myosin. Competition and molecular docking studies suggest that BHC and blebbistatin bind to the same site on myosin.

Full text of DOI:10.1021/acs.jmedchem.0c01062

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Article
Synthesis and Evaluation of 4-Hydroxycoumarin
Imines as Inhibitors of Class II Myosins
Jhonnathan Brawley, Emily Etter, Dante Heredia, Amarawan Intasiri, Kyle Nennecker, Joshua D. Smith,
Brandon Welcome, Richard K. Brizendine, Thomas W. Gould, Thomas W. Bell, and Christine R. Cremo
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c01062 • Publication Date (Web): 07 Sep 2020
Downloaded from pubs.acs.org on September 7, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis and Evaluation of 4-Hydroxycoumarin Imines as
Inhibitors of Class II Myosins
†
‡
§
†,
Jhonnathan Brawley, Emily Etter, Dante Heredia, Amarawan Intasiri, Kyle
‡
‡
‡
‡
Nennecker, Joshua Smith, Brandon M. Welcome, Richard K. Brizendine ,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
,
§
,†
‡
Thomas Gould,* Thomas W. Bell,* and Christine Cremo*
†Department of Chemistry, University of Nevada, Reno, Nevada 89557-0216 USA
‡Department of Pharmacology, University of Nevada, Reno School of Medicine, Reno,
Nevada 89557-0318 USA
§Department of Physiology and Cell Biology, University of Nevada, Reno School of
Medicine, Reno, Nevada 89557-0352 USA
Department of Chemistry, Faculty of Science, Chulalongkorn University, Bangkok, 10330
Thailand
ABSTRACT: Inhibitors of muscle myosin ATPases are needed to treat conditions that could be
improved by promoting muscle relaxation. The lead compound for this study (3-(N-
butylethanimidoyl)ethyl)-4-hydroxy-2H-chromen-2-one; BHC) was previously discovered to
inhibit skeletal myosin II. BHC and 34 analogs were synthesized to explore structure-activity
relationships. Properties of analogs, including solubility, stability and toxicity, suggest that the
BHC scaffold may be useful for developing therapeutics. Inhibition of actin-activated ATPase
activity of fast skeletal and cardiac muscle myosin II, inhibition of skeletal muscle contractility ex
vivo, and slowing of in vitro actin-sliding velocity were measured. Several analogs with aromatic
side arms showed improved potency (IC < 1 µM) and selectivity (≥ 12-fold) for skeletal myosin
5
0
versus cardiac myosin compared to BHC. Several analogs blocked neurotransmission, suggesting
they are selective for nonmuscle myosin II over skeletal myosin. Competition and molecular
docking studies suggest that BHC and blebbistatin bind to the same site on myosin.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
Myosins are essential actin-activated motor proteins that hydrolyze adenosine triphosphate
1
(ATP) to produce motion and force under load. Class II myosins are found in the contractile
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
apparatus and are responsible for voluntary and involuntary muscle contraction. Distinct isoforms
2
are found in skeletal, cardiac, and smooth muscles. Differences in kinetics of the ATPase cycle
(Figure 1) tune physiological parameters specific to muscle function such as contractile speed,
3
maximal force, and force maintenance. There are also non-muscle myosin II isoforms that are
found in stress fibers of nonmuscle cells. They mediate diverse cellular processes such as cell
division, migration, and adhesion. Class II myosins share a conserved motor domain containing
the actin-binding and ATP-binding sites, a lever arm domain to amplify motion, and a tail domain
to allow filaments to form through intermolecular interactions (Figure 1).4
Under activating cellular conditions, myosin cycles between actin-attached and actin-detached
states, controlled by the interaction of ATP with myosin, which weakens actin binding affinity
(Figure 1). After conformational changes triggered by ATP hydrolysis, myosin weakly binds to
actin and upon release of inorganic phosphate (P ) transitions to a strongly bound adenosine
i
diphosphate (ADP) state, which undergoes the working step, resulting in relative motion between
actin and myosin or in the generation of force.3
Several clinical conditions are characterized by excessive muscle tone or activity. Examples
5
6
include spasticity of skeletal muscle as a result of genetic mutations or spinal cord injuries and
7
cardiomyopathies caused by mutations in cardiac sarcomeric proteins including myosin. Several
conditions are caused by excessive bronchial, gastrointestinal, and urinary smooth muscle
8
9
contractility, such as asthma, diarrhea and overactive bladder. In some of these conditions,
ACS Paragon Plus Environment
2

Page 3 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
genetic alterations directly lead to dysregulated myosin activity,10 whereas in others, excess
myosin activity is an indirect consequence of neural, immune or other dysfunction.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Myosin ATPase cycle. Myosin-ATP and -ADP-P (blue) bind weakly to actin and are
i
shown in actin-detached states. All other myosin states (red) are shown in the actin-attached state
along with their relative affinities for actin. The overall rate of attachment of myosin to actin, katt,
is limited by initial weak binding of the ADP-P state (K ) followed by a transition to a strongly
i
w
bound ADP state coupled to P release (k ). A conformational change swings the lever arm
i
ws
domain resulting in the working step. After ADP release, ATP rebinds, which leads to weakly
bound states. Image modified from Ref 4.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Structures of blebbistatin (1a), para-aminoblebbistatin (1b), and BHC (2). The coumarin
ring in 2 is inverted relative to the usual representation to show proposed pharmacophore,
consisting of aromatic core, hydrogen bonding and hydrophobic side chain.
As a result of this central role in mediating muscle contractility, myosins have emerged as an
appealing target for small molecule inhibitors aimed at ameliorating muscle hypercontractility.11,
1
2
However, myosin II is a difficult target due to its presence in three different kinds of muscle as
well as nonmuscle cells. To promote muscle relaxation, inhibitors must stabilize myosin states that
interact weakly with actin (Figure 1). Kinetically, this means that the rate limiting step in the
ATPase cycle, k , must be slowed, trapping myosin in the ADP-P state. One of the best-
att
i
1
3
characterized inhibitors, blebbistatin (1a, Figure 2, review ), inhibits k for skeletal, cardiac, and
att
vertebrate non-muscle myosins in the low micromolar range, but is less potent for smooth muscle
1
1
myosin. Photoinstability, phototoxicity, and low solubility of blebbistatin have led to searches
for analogs with more favorable physicochemical properties,14-19 such as the less potent para-
aminoblebbistatin (1b) with improved photostability, solubility, and phototoxicity.14, 15, 19 The
2
0
most selective myosin inhibitor known (CN-2018571) targets smooth muscle myosin II and
binds to a different site than blebbistatin.
ACS Paragon Plus Environment
4

Page 5 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The objective of this work was to synthesize and characterize novel, small molecule myosin
inhibitors as potential therapeutics for patients with symptoms of diseases or syndromes that could
be improved by promoting muscle relaxation. The results of a small-molecule library screen21
showed that the compound 3-(N-butylethanimidoyl)-4-hydroxy-2H-chromen-2-one (BHC, 2,
Figure 2) exhibited a detectable but unspecified decrease in the actin-activated ATPase activity of
skeletal muscle myosin at 100 µM. However, no further properties were investigated, including
the half-maximal inhibitory concentration (IC ), mechanism of action, or selectivity toward
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
different muscle myosin isoforms. Subsequently, we demonstrated that 2 selectively blocks nerve-
2
2
evoked contraction of mouse skeletal muscle without affecting neurotransmitter release, which
is required for the action potential. This established a new tool to directly evaluate
neurotransmission by measuring successfully transduced muscle action potentials without
interference by muscle movement, which damages electrodes and prevents accurate readings.
These useful properties of 2 were not found for the myosin inhibitor N-benzyl-p-
21
22
toluenesulfonamide (BTS), which blocked neurotransmission in addition to muscle movement.
Structural similarities between blebbistatin (1a) and BHC (2) shown in Figure 2 suggest that
they share the same pharmacophore and bind to the same site of myosin II. In this study, we
synthesized and characterized 2 and 34 analogs, then examined them for their ability to inhibit
myosin function using two in vitro assays and one ex vivo muscle contractility assay. We have
determined the IC values for inhibition of actin-activated ATPase activity for both fast skeletal
50
and cardiac muscle myosin. From in silico molecular modeling and by competition assays we
provide evidence that 2 and an analog of 2 bind to the same site on myosin as blebbistatin (1a).
Interestingly, several analogs exhibit higher affinity and are more selective for skeletal or cardiac
muscle myosin than either 2 or 1a, which are of comparable potency. As predicted by our prior
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
kinetic models,23-25 2 slows in vitro actin sliding velocity with IC values that are similar to those
5
0
for inhibition of actin-activated ATPase, suggesting a common kinetic mechanism of action for
both effects. Finally, by combining information from in vitro and ex vivo assays, several analogs
appear more selective for nonmuscle myosin than either skeletal or cardiac myosin.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Scheme 1
Product
Yield
78%
R’
Product
Yield
72%
R
XAr
Bn
2
(BHC)
Bu
7a (JB002)
Me
6
6
6
6
6
a (JB032)
b (JB010)
c (JB009)
d (JB011)
e (JB003)
40%
59%
53%
73%
57%
67%
49%
65%
91%
83%
75%
H
Me
7b (JB070)
7c (JB069)
7d (JB060)
7e (JB063)
7f J(B076)
7g (JB029)
7h (JB030)
7i (JB031)
8a (JB064)
8b (JB066)
72%
82%
65%
65%
94%
78%
75%
70%
91%
93%
Me
Me
Me
Me
Me
Me
Me
Me
Et
CH
CH
CH
2
2
2
-2-MePh
-3-MePh
-4-MePh
Et
Pr
OCH
2
Ph
i-Pr
i-Bu
sec-Bu
NHPh
6
f (JB004)
CH
2
2
2
-2-Py
6
g (JB012)
h (JB005)
CH
CH
-3-Py
-4-Py
6
(CH
(CH
2
2
)
4
5
CH
CH
3
3
6
i (JB052)
j (JB006)
k (JB007)
)
Bn
Bn
6
(CH
(CH
2
)
2
OH
OH
Pr
6
2 3
)
ACS Paragon Plus Environment
6

Page 7 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis. Previous studies on blebbistatin analogs have found that variation of the
hydrophobic side chain region can tune relative inhibitory potency for myosins. According to the
proposed pharmacophore comparison shown in Figure 2, this corresponds to the butylimine side
chain of BHC (2), so this became the initial focus of analog studies. Preparation of the Schiff base
BHC (2, Figure 1) by condensation of butylamine with 3-acetyl-4-hydroxycoumarin (3-acetyl-4-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6
27
hydroxy-2H-chromen-2-one, 4, Scheme 1) was previously reported. Bicyclic compound 4 is a
scaffold for several biologically active compounds and is readily prepared from inexpensive 4-
hydroxycoumarin (4-hydroxy-2H-chromen-2-one, 3) by base-catalyzed acetylation with acetyl
2
8
29
chloride or by reaction with acetic acid or acetic anhydride and phosphorus oxychloride. As
shown in Scheme 1, the latter method was used to prepare 4, as well as the 3-propionyl (5a) and
3
-butanoyl (5b) analogs. Given the ease of condensing these ketones with various commercially
3
0
31
available amines or their hydrochloride salts, a series of 20 BHC analogs (6a–k and 7a–i) were
prepared from 4 with various imine side chains to explore SAR. The N-benzyl analog 7a was found
early in our studies to have promising inhibitory activity, so we also prepared analogs 8a and 8b
replacing the acetyl methyl group with bulkier ethyl and propyl groups, respectively. Most of these
compounds were isolated in high yields after recrystallization to obtain high purity. Another hit
2
1
from the same small-molecule screen as BHC (2), 5-nitrocoumarin 10a bearing a pentyl side
3
2
chain, was synthesized as shown in Scheme 2 via nitration of 3-acetyl-4-hydroxy-2H-chromen-
-one (4), followed by condensation with pentylamine. Since the N-benzyl analog 7a lacking a
2
nitro substituent showed activity in initial studies, intermediate 9 was also condensed with
benzylamine to give compound 10b. Nitro groups are often associated with toxicity issues in
3
3
drugs, so further nitro analogs were not pursued.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
34
Quinoline and quinolinone ring systems are favorable scaffolds in drug candidates, so a series of
4
-hydroxyquinolin-2-one analogs was synthesized (Scheme 3). Replacing the 1-position oxygen
atom of BHC with nitrogen should increase the dipole moment of the ring system and also increase
the hydrogen bond acceptor (HBA) ability of the 2-position carbonyl oxygen. Key intermediate,
3
-acetyl-4-hydroxyquinolin-2(1H)-one (11), was prepared from inexpensive starting materials
3
5
ethyl acetoacetate and methyl anthranilate. The low potency of 12a-i (vide infra) suggested that
replacing the oxygen in the coumarin series (2, 6 and 7) with an NH group may perturb binding
by introducing a hydrogen bond donor (HBD) in the 1-position of the ring system. Accordingly,
compound 14, an N-methyl analog of 2 and 12c, was prepared as shown in Scheme 3.
Structure. Schiff bases derived from 3-acyl-4-hydroxycoumarins or 3-acyl-4-
hydroxyquinolin-2-ones can exist in either of two tautomeric forms, labeled “imine” and
“
enamine” in Figure 3. While both can be stabilized by intramolecular hydrogen bonding, X-ray
crystal structures of 6j and 7a show only the enamine isomers,36-38 and the NH–O hydrogen bond
in the enamine form of 7a has been studied computationally.36, 38, 39 In addition, the structures of
1
13
6
a, 6j, and 7a were assigned as enamine tautomers from their H and C NMR spectra in CDCl
3
4
0
and DMSO-d6.
ACS Paragon Plus Environment
8

Page 9 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 3
Product
2a (JB038)
12b (JB039)
Yield
63%
52%
17%
78%
56%
11%
70%
72%
5%
R
Pr
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
i-Pr
12c (JB033)
12d (JB040)
12e (JB041)
Bu
i-Bu
sec-Bu
1
2f (JB037)
12g (JB047)
2h (JB043)
2i (JB036)
(CH
(CH
CH
2
)
4
Me
2
)
2
OH
1
2
Cy
1
Bn
Figure 3. Imine and enamine tautomers of 3-acyl-4-hydroxycoumarins (X = O) and 3-acyl-4-
hydroxyquinolin-2-ones (X = NR).
O-Benzyloxime 7e and phenylhydrazine 7f should exist primarily as imine tautomers under
most conditions, but the prevalent isomer of 2, 6a-k, 7a-d, 7g-i, 10a,b, 8a,b, 12a-i, and 14 in
solution could depend on solvent characteristics. In CDCl solution, a single peak at    ppm
3
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
13
for the ketone carbonyl carbon is observed in the C NMR spectra for the enamine structures of
6
a, 6c, 6d, and 6g. As in the crystal structures, intramolecular hydrogen bonding should stabilize
the enamine tautomers in chloroform. DMSO is a strong HBA and could break the intramolecular
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hydrogen bond, but in DMSO-d compounds 2, 6a-k, 7a-d, 7g-i, 10a,b, 8a,b, 12a-i, and 14
6
1
3
consistently show a ketone carbonyl resonance at  ~ 180 ppm in the C NMR spectra and HNCH
1
splitting is observed in the H NMR spectra. To maintain consistency with nomenclature in the
previous literature, these compounds are named as imine tautomers in the Experimental section,
1
but the assignments for the H NMR spectra are based on the enamine structures. Consistent with
the expectation that O-benzyloxime 7e and phenylhydrazine 7f would exist primarily as the imine
tautomer, their NMR spectra showed no ketone carbonyl resonance and no HNCH splitting. These
observations are relevant to computational docking studies because the locations of HBD and HBA
groups on small molecules affect specific interactions with amino acids forming the binding site.
Structure-activity relationships. To evaluate the relative ability of compounds to inhibit
myosin II function, we measured the IC values for reduction of steady-state actin-activated
5
0
2
+
Mg ATPase activity for skeletal and cardiac myosin. This assay measures the rate-limiting step
in the acto-myosin ATPase cycle (k , Figure 1), which is the kinetic parameter most related to the
att
ability to inhibit muscle contraction. We chose to use an assay that directly measures production
of P with time, detecting absorbance in a stopped-time format. This avoids potential interference
i
by light scattering or fluorescence. Figure 4A shows representative plots of ATPase activity versus
concentration of selected BHC analogs.
ACS Paragon Plus Environment
1
0

Page 11 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. A. Representative titrations of analogs versus ATPase activity of skeletal myosin. BHC
(2) red circles, 7i, black squares and 6b, blue diamonds. Lines show fits as described in text. Error
bars show standard deviation for N determinations (see Table 1). B. Linear aliphatic chain length
of BHC (2) analogs from Table 1 versus IC for skeletal (blue) and cardiac (red) myosin. Points
5
0
are connected by lines. Under these conditions KATPase = 8.7 ± 1.5 μM and Vmax ranges from 0.9
-1 -1
to 1.5 s h . Conditions: 30 °C, assay buffer with 32 µM actin, 2 mM ATP.
Table 1 contains all IC data obtained from a fit to a four-parameter logistics equation as well
50
as the number of replicates (N), i.e. the number of times a fully independent titration/ATPase assay
was performed. The IC value for 1a was within the range of published values for skeletal myosin
5
0
1
5
(
0.01 – 4 µM) and cardiac myosin (0.4 – 30 µM), respectively. Values for 1b were similarly
1
4
within range. Both compounds were more potent inhibitors of skeletal versus cardiac myosin by
about a factor of 4.5.
Table 1B shows data for compounds with linear aliphatic groups appended to the imine
nitrogen. BHC (2) was found to be a less potent inhibitor than 1a and 1b by about a factor of 3.
ACS Paragon Plus Environment
1
1

Journal of Medicinal Chemistry
Page 12 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
However, it fits the same trend as the blebbistatins, being ~ 3-fold more potent for skeletal versus
cardiac myosin. Increasing the aliphatic chain length did not increase this selectivity but had a
significant effect on potency for both skeletal and cardiac myosin (Figure 4b). Compounds having
side arms with fewer than four carbons were poor inhibitors. Increasing chain length from four to
six carbons gave at most (6i) ~ 2-fold increase in potency over 2. Appending a terminal OH group
had little effect on potency of selectivity. Branching at the α- or -carbon of the side arm did not
improve inhibitory potency of BHC analogs with less than 4 carbons (Table 1C).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
+
Table 1A-G Potencies (IC Values) for Inhibition of Steady-State Actin-Activated Mg
5
0
ATPase Activity of Skeletal and Cardiac Myosin.
a
a
IC50 (M)
IC50 ratio:
cardiac/
skeletal
IC50 (M)
IC50 ratio:
cardiac/
skeletal
(Number replicates, N)b
(Number replicates, N)b
Compound
Structure
Compound
Structure
Cardiac
Skeletal
Cardiac
Skeletal
Table 1A: Blebbistatins
Table 1D: BHC analogs with benzene rings (continued)
O
OH
Me
2.3 ± 0.3
0.5 ± 0.1
(1)
7e

100
100
O
OH
O
N
O
1a
N
N
4.6
4.2
(1)
(JB063)
(1)
(1)
O
OH
Me
H
2
.1 ± 0.7
1)
0.5 ± 0.7
(1)
7f
N
N
27 ± 13
(3)
2.0 ± 1.0
(3)
OH
O
N
N
1b
13.5
(
(JB076)
O
NH2
8a
OH
O
N
O
100
(1)
100
(1)
Table 1B: BHC and analogs with linear aliphatic chains
(
JB064)
2
OH
O
N
O
5.9 ± 1.2
(3)
2.1 ± 0.8
(3)
8b
OH
O
N
O
100
(1)
100
(1)
2.8
BHC
(JB066)
OH NH
6
a
100
100
Table 1E: BHC analogs with pyridylmethyl groups
N
(
JB032)
O
O
(1)
(1)
7g
OH
N
22 ± 11
(3)
6
.8 ± 1.3
3)
3
5
.2
.6
(
(
JB029)
O
O
OH
N
6b
100
100
(JB010)
(1)
(2)
O
O
7
h
OH
O
N
O
N
17 ± 1
(3)
3.0 ± 0.9
3)
6
c
OH
N
O
100 ± 50
(3)
47 ± 9
(3)
(
(
JB030)
2
1
.1
.7
(
JB009)
O
OH
O
6
d
N
O
54 ± 14
(3)
31 ± 3
(4)
7i
OH
O
N
O
8.5 ± 0.8
(3)
N
100
12
(
JB011)
(JB031)
ACS Paragon Plus Environment
1
2

Page 13 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(3)
Table 1F: BHC analogs with quinolinone ring system
6h
OH
O
N
O
6.7 ± 0.8
(3)
2.0 ± 0.4
(3)
OH
N
3
2
.3
.4
12a
100
100
(
JB005)
N
O
(
JB038)
H
(1)
(1)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
i
OH
O
N
2.9 ± 0.9
(3)
1.2 ± 1.0
(3)
12b
OH
N
O
100
100
O
(JB052)
(JB039)
(1)
(1)
N
H
6j
OH
N
O
OH
92 ± 25
(1)
92 ± 52
(2)
12c
OH
N
O
100
(1)
100
(1)
1
1
.0
.2
N
(
JB006)
O
(JB033)
12d
H
OH
N
O
6
k
OH
O
N
OH
37 ± 7
(3)
31 ± 4
(3)
100
(1)
100
(1)
O
N
H
(JB007)
(JB040)
12e
OH
N
O
100
100
Table 1C: BHC analogs with branched-chain aliphatic groups
(
JB041)
N
H
(1)
(1)
6
e
100
100
12f
OH
N
100
100
OH
O
N
O
N
H
O
(
JB003)
(3)
(2)
(JB037)
(1)
(1)
OH
6
f
OH
N
O
34 ± 4
(4)
 100
12g
OH
N
100
100

0.34
N
H
O
(JB004)
O
(3)
(JB047)
(1)
(1)
6
g
100
100
12h
OH
N
O
100
100
OH
O
N
O
N
H
(
JB012)
(1)
(1)
(JB043)
(1)
(1)
1
2i
OH
N
O
100
(1)
100
(1)
Table 1D: BHC analogs with benzene rings
N
H
(JB036)
14
OH
N
N
O
7
a
OH
O
N
O
7.3 ± 0.7
(3)
0.8 ± 0.2
(3)
100
(1)
100
(1)
9
.1
(
(
(
JB002)
(JB075)
7b
100
(2)
100
(1)
OH
O
N
O
Table 1G: Nitrocoumarin BHC analogs
JB070)
1
0a
100
100
7
c
OH
O
N
4.9 ± 1.7
(3)
10 ± 4
(3)
(
JB096)
(1)
(1)
0.5
O
JB069)
7
d
OH
O
N
O
7.9 ± 2.5
(3)
0.8 ± 0.3
(4)
10b
100
100
9
.9
(JB060)
(JB065)
(1)
(1)
a
Uncertainty (±) is the standard error of the fit to the combined data from N independent titration experiments (see text).
ACS Paragon Plus Environment
1
3

Journal of Medicinal Chemistry
Page 14 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Benzyl (7a) and 4-methylbenzyl (7d) analogs of 2 showed a ~10-fold higher potency for
skeletal than for cardiac myosin and a ~ 2-fold improvement in potency over 2 for skeletal myosin
(Table 1D). Both of these effects were absent in the 3-methylbenzyl analog 7c, and the 2-
methylbenzyl analog 7b, which lacked significant potency. Because of this sensitivity to
modification, further derivatives of 7a were prepared to examine the effect of lengthening the
imine side arm, but this caused a loss of potency as in 8a and 8b. Derivatives with insertion of a
heteroatom (7e and 7f) did not lead to improvements in potency relative to 7a, although 7f retained
high selectivity for skeletal myosin. Notably, isobutylimine 6f and m-methylbenzylamine 7c
displayed some degree of selectivity for cardiac vs. skeletal myosin, suggesting the potential to
develop drugs for cardiomyopathies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The pyridine analogs, 7g, 7h, 7i, (Table 1E) all showed a similar trend toward decreased
potency relative to 7a but largely for cardiac myosin. The 4-pyridylmethyl analog 7i and
phenylhydrazine 7f showed highest selectivity of any derivatives in Table 1 with ≥ 12-fold
preference for skeletal vs. cardiac myosin.
None of the quinolinone derivatives, 12a-12i, 14 (Table 1F), had measurable potency toward
cardiac or skeletal myosin, showing that even with favorable imine side arms, the oxygen in the
1
-position is critical to inhibiting ATPase activity and cannot be replaced with a NH or N-methyl
as in 14. Similarly, BHC analogs with a 6-nitro group (Table 1G) 10a and 10b lacked potency;
0b was previously found to have detectable but unspecified inhibitory activity for skeletal myosin
1
2
1
(at 100 µM) in the small molecule screen that uncovered BHC, but we find that it is a poor
inhibitor for both skeletal and cardiac myosins (IC  100 µM), in agreement with our previous
5
0
demonstration of its inability to paralyze skeletal muscle tissue.4, 22
ACS Paragon Plus Environment
1
4

Page 15 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Effects of BHC on velocity of actin sliding in an in vitro motility assay. Another measure
of myosin function can be assessed using the in vitro motility or actin-sliding assay in which
fluorescent actin filaments are observed moving over a bed of myosin attached to a coverslip.41
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
Figure 5. Effect of BHC (2) on in vitro actin sliding velocity of skeletal myosin at 100 µg mL .
Error bars show standard deviations and red line shows fit giving IC = 2.8 ± 0.4 µM (± is standard
50
deviation of error of the fit). T = 30 °C.
Figure 5 shows that BHC inhibits actin sliding velocities with an IC of 2.8 ± 0.4 µM, a value
5
0
very close to the IC for the ATPase data (2.1 ± 0.8 µM). This finding is predicted by our work
5
0
showing that velocity is not solely detachment-limited in this assay but rather also depends on the
acto-myosin attachment rate, which is limited by the ATPase activity.23-25
Mechanism of interaction with myosin. Since blebbistatin (1a) and potent BHC
analogs inhibit ATPase activity and in vitro motility, we determined whether they bind
ACS Paragon Plus Environment
1
5

Journal of Medicinal Chemistry
Page 16 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
to the same site on myosin. In search for a simple method to measure binding, we
discovered that 1a shows a ~ 40-fold increase in fluorescence emission at 560 nm (λ 440
ex
nm) upon saturation with the soluble head domain (S1) of skeletal myosin, but only in
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the presence of 1 mM ATP (Figure 6A). This is consistent with kinetic studies showing
42
that 1a has high affinity only for the myosin-ADP-P transition state. Without ATP, 1a
i
fluorescence remained at background levels (data not shown). S1 was added to a fixed
concentration of 1a and data were fit to the quadratic form of the equation for a single-
site binding equilibrium to give K for 1a = 0.4 ± 0.1 µM (Figure 6A). This value is
d
essentially the same as the IC of 1a for skeletal myosin (Table 1).
5
0
Figure 6. Interactions of 1a and 7a with skeletal S1 at 22-25 °C in assay buffer with 1 mM
ATP. A. Fluorescence of a mixture of 1a (2 µM) and S1 was measured at 560 nm (λ 440
ex
nm). B. Fluorescence of a mixture of 1a (2 µM) and S1 (1 µM) was measured after
ACS Paragon Plus Environment
1
6

Page 17 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
incubating with 7a for 1 hr. Fluorescence without S1 (F ) was subtracted from
-S1
fluorescence with S1 (F ) and normalized to maximal fluorescence (Fmax). Curve fitting
+
S1
(red lines) is described in Experimental section. N = 1.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound 7a was selected as a representative compound to determine whether BHC analogs
can compete with 1a for binding to myosin. In contrast to 1a, the weak fluorescence of 7a (λ 320
ex
nm, λ 400 nm) did not change upon adding S1 (with MgATP; data not shown). Figure 6B shows
em
that adding increasing 7a to a mixture of S1 and 1a caused the fluorescence of 1a to decrease,
suggesting that 7a and 1a compete for the same binding site on S1. A fit to a four-parameter
logistics equation gives an estimate the IC value of 7a for S1 of 5.3 ± 1.1 µM with a slope = 0.8.
5
0
4
3
The K (apparent K ) was calculated from the IC value using the equations described or by
i
d
50
using the IC -to-K converter (https://bioinfo-abcc.ncifcrf.gov/IC _Ki_Converter). Free 7a
5
0
i
50
concentrations were applied to a competitive binding model, giving K = 0.8 ± 0.2 µM, which is
i
essentially the same value as the IC for inhibition of the ATPase activity (Table 1D).
5
0
These data suggest that 1a and 7a compete for the same binding site on S1 with similar
affinity.
Computational Modeling. The data shown in Figure 6 suggest that 7a and
blebbistatin (1a) bind to the same site in skeletal myosin. The atomic resolution structure
4
4
of the motor domain of Dictyostelium discoideum myosin (PDB 1yv3 ) shows that 1a binds
to a hydrophobic pocket at the apex of the open conformation of the cleft between the
upper 50 and lower 50 kDa subdomains of the motor domain (Figure 7C). The cleft must
ACS Paragon Plus Environment
1
7

Journal of Medicinal Chemistry
Page 18 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
5
close to allow strong binding to actin, explaining why 1a inhibits the release of P from
i
the ADP-P complex, which is normally accelerated by actin binding (Figure 1).
i
To evaluate structural aspects of analog binding, we prepared two homology models
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
cardiac and skeletal myosin) using 1yv3 as the template and compared in situ docking to the
models and template. The highest scoring pose from docking 1a to 1yv3 was essentially the
same as the known structure (data not shown).
Docking to the cardiac structure showed that 1a (Figure 7A) and 2 (Figure 7B) bound to the
same site as 1a in 1yv3. Table 2 summarizes contact residues and interactions found from docking
1
a and 2, which share a majority of residue interactions. Ala463, located directly adjacent to ATP
on Switch 2 next to P in the ADP-P transition state, as well as Gly245, and Leu267 form hydrogen
i
i
bonding interactions with 1a (Figure 7A). Two of these residues also participate in polar
interactions with 2, specifically the amide nitrogen of Ala463 and amide oxygen of Leu267 (Figure
7
B). Molecule 1a forms -stacking interactions between the A-ring and Tyr266 with parallel
aromatic planes at 4.2 Å, as well as an edge-to-face  interaction between the D-ring and Phe473,
with perpendicular aromatic rings at 4.9 Å. These distances are within the range of known -
4
6
interactions. Comparable - interactions are also formed between the aromatic coumarin core
of 2 and the phenyl ring of Tyr266, at the same distance as the interaction between Tyr266 and 1a
(Figure 7A).
ACS Paragon Plus Environment
1
8

Page 19 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Docking results. A. Highest scoring pose of 1a (green), and B. highest scoring
pose of 2 (yellow) docked to cardiac myosin homology model (grey). Residues
participating in ligand binding are shown as light blue sticks and polar interactions are
shown as red dashed lines. C. Structure of D. discoideum myosin II (PDB:1yv3, grey)
showing 1a (green) and ADP-VO4 (red). D. Residues (light blue) within 4 Å of the phenyl
group of 7a (orange) in cardiac myosin model. Non-carbon atoms for ligands: red for
oxygen, blue for nitrogen, and white for amide hydrogens.
ACS Paragon Plus Environment
1
9

Journal of Medicinal Chemistry
Page 20 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. Docking results listing contact residues within 4Å of the highest scoring
binding pose of 1a and 2, with bolded residues and ligand functional groups involved
in hydrogen bonding.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
BHC (2)
Blebbistatin (1a)
Residue in
bovine
cardiac
myosina
Leu658
Asn654
Tyr266
His651
Thr481
Ala463
Gly245
Leu267
Ile478
O
OH
Me
Residue in
rabbit skeletal
myosinb
O
HN
Residue
in 1yv3
4
A
B
C
3
N
N
A
B
2
O
1
O
D
A-ring
A-ring
A-ring
A-ring
A-ring
B-ring
B-ring
Leu641
Gln637
Tyr261
Tyr634
Thr481
Ser456
Gly240
Leu655
Asn658
Tyr268
Phe655
Thr483
Ala465
Gly247
Leu269
Ile480
A-ring
4-carbonyl
B-ring
2-carbonyl
B-ring
carbonyl
hydroxyl oxygen
enamine nitrogen hydroxyl hydrogen Leu262
enamine nitrogen
no interaction
no interaction
C-ring
D-ring
D-ring
Ile471
Glu467
Phe466
Glu474
Phe473
Glu476
Leu475
a
b
Gene: MYH7; UniprotKB: Q9BE39 (MYH7_BOVIN). Gene: MYH13; UniProtKB:
Q9GJP9 (Q9GJP9_RABIT)
ACS Paragon Plus Environment
2
0

Page 21 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Docking with both the imine and enamine tautomers of 2 (Figure 3) suggests that the enamine
has the most favorable interactions. The highest scoring pose for the enamine shows the carbonyl
oxygen at the 4-position near a HBD group, the protonated imidazole ring of His651. The carbonyl
oxygen in the enamine is a more favorable HBA than the hydroxyl group at this position in the
imine. The nitrogen of 2 was also near an HBA, the amide oxygen of Leu267 (Figure 7B). Also,
the planar conformation of the enamine better fits the shape of the pocket. The enamine resulted
in higher scoring poses than the imine, which is consistent with the prevalence of enamine
tautomers for BHC compounds, as shown by NMR spectra as previously discussed.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The region of the pocket surrounding the variable imine side arm of BHC analogs is
hydrophobic, consisting of aliphatic regions of Val647, Phe473, Glu474, Lys270, Glu269, Leu267
in cardiac myosin (Figure 7D). Similar to the N-phenyl group of 1a, the face of the phenyl group
of 7a is at a near-right angle to the aromatic plane of Phe473. This capacity for edge-to-face
aromatic interactions could explain differential affinity for certain BHC analogs having aromatic
rings in their side arms, but does not explain selectivity for skeletal versus cardiac myosin (e.g.
7
a). Longer aliphatic or aromatic rings are predicted to have more potential for hydrophobic
interactions, generally consistent with correlation between increasing aliphatic side chain length
and greater potency for ATPase inhibition (Figure 4B). These residues surrounding the side arm
are mostly conserved, except a leucine residue in skeletal myosin is at the position of Phe473 in
cardiac myosin (Table 2). The only other residue difference between the binding pockets is
replacement of the His651 residue in cardiac by phenylalanine in skeletal myosin.
Docking experiments using the two homology models showed that 1a and most of the analogs
(enamine and imine tautomers) docked to the 1a binding site in the highest scoring pose.
Interesting exceptions were 6h, 6i, 10a, 10b, 8a and 8b, which each docked to unique locations.
ACS Paragon Plus Environment
2
1

Journal of Medicinal Chemistry
Page 22 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Inspections of the structures (Table 1) show that they contain either long side arms (6h (5 C), 6i
(6 C)), nitro groups on the coumarin ring (10a, 10b (5 C)), or an extended side arm attached to the
imine carbon (8a (2 C), 8b (3 C)). Of these analogs, only 6h and 6i have good potency by ATPase
assay. Lack of docking to the 1a binding site may suggest that the site must rearrange to
accommodate the long side arm.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Physiological studies. Our biochemical and modeling results suggest that BHC analogs and
1
a bind to the same site (Figures 6 and 7). ATPase data show that some compounds are selective
for skeletal vs. cardiac myosin (like 1a), have opposite selectivity (e.g., 6f and 7c, Table 1), or lack
selectivity. To test whether analogs exhibit additional selectivity for the predominant Class II
myosins found in nerves, nonmuscle myosin IIB, as well as to examine their effects in ex vivo
muscle contractility assays, we took advantage of the mouse phrenic nerve-hemidiaphragm
preparation. The hemidiaphragm can be stimulated indirectly through electrical pulses delivered
through a suction electrode attached to the phrenic nerve. This allows for simultaneous evaluation
of the effects of compounds on skeletal muscle contractility and on neurotransmission, which
depends on non-muscle myosin IIB.47
ACS Paragon Plus Environment
2
2

Page 23 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 8. Representative effect of BHC analogues on phrenic nerve stimulation-induced
contraction of mouse diaphragm. Muscle force represented by optical shortening of diaphragm
fibers in response to nerve stimulation at 1Hz for 5s (blue, vehicle; red, 100 µM 7i).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We measured contractility optically as diaphragm muscle fiber shortening in response to
2
2
stimulation of the phrenic nerve in the presence of selected compounds (100-300 M). Figure 8
shows a representative example of contractility data of muscle treated with 7i, which essentially
blocked muscle contraction at 100 µM. Next, in the case where analogs produced >85% block in
contractility at any dose, we attempted to record changes of transmembrane membrane potential
in muscle in response to phrenic nerve stimulation in the presence or absence of the compounds.
We reasoned that if a drug selectively blocked skeletal muscle myosin but not non-muscle myosin
IIB, we should be able to record muscle action potentials, since these electrical potentials depend
on the activity of voltage-gated sodium channels upstream of calcium release and subsequent actin-
myosin interactions.
Compound 7i effectively paralyzed muscle at 100 M without blocking neurotransmission
2
2
(
Table 3), as we have previously shown for 2. Thus, 2 and 7i form Class 1 in Table 3, which
inhibit skeletal muscle but not nerve nonmuscle myosin IIB. This important selectivity has not
been demonstrated to date for other known myosin inhibitors to our knowledge. Compared to 2,
7
i has a much higher potency for skeletal than cardiac myosin as well, an attractive property for
treating skeletal hypercontractility.
Class 2, which includes 1a, effectively paralyzed diaphragm muscle at 100 M, generally
consistent with the potency assessed by ATPase activity. However, action potentials were also
4
8
blocked (Table 3), consistent with the inhibitory effects of 1a on nonmuscle IIB as well as on
ACS Paragon Plus Environment
2
3

Journal of Medicinal Chemistry
Page 24 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
47
synaptic transmission previously reported. It is likely that these compounds have potency for
both skeletal myosin and nonmuscle myosins in the nerve terminal.
Class 3 analogs are similar to Class 2 with respect to observed muscle effects, but have poor
potency with respect to ATPase activity. This reduction in contractility is therefore likely a
consequence of the upstream effect to block nerve activity, suggesting that Class 3 analogs are
likely selective for nonmuscle myosin IIB over skeletal myosin. Interestingly, all compounds in
this class are also poor inhibitors of cardiac myosin (Table 1, IC  100) and therefore appear to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
be relatively selective for nonmuscle IIB. Among these are five quinolinones (12d-e, 12f-h), but
not all quinolinones tested have this property (12a-c, 12i). Among the coumarin analogs found to
be in Class 3 (6a-b, 6e), a common feature is a short side arm consisting of 0, 1, or 3 carbons,
respectively.
Finally, analogs in Class 4 only weakly reduced nerve stimulation-induced skeletal muscle
contractility or actin-activated ATPase activity of skeletal myosin, and therefore are poor inhibitors
of skeletal and nonmuscle myosin IIB. Among them, only 6f has an IC value of <100 µM for
5
0
cardiac myosin (Table 1) suggesting that it may be a good lead compound for cardiac selectivity.
Table 3. Summary of effects of compounds on skeletal muscle contractility and
neurotransmission.
Decrease in muscle
shortening (%) at 100 M
(N, number of replicates)
Action potentials
observed at paralyzing
doses?
Skeletal myosin
Compound
ATPase IC50
100 M?
<
Class 1: Potency for skeletal but not nerve myosin
2
(BHC)
96 ± 2 (5)
98 ± 3 (2)
Yes
Yes
Yes
Yes
7
i (JB031)
ACS Paragon Plus Environment
2
4

Page 25 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Class 2: Potency for both nerve and skeletal myosin
1
a (blebbistatin)
j (JB006)
k (JB007)
g (JB029)
99 (1)
87 (1)
No
No
Noa
No
Yes
Yes
Yes
Yes
6
6
72 ± 3 (2)
86 (1)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
Class 3: Higher potency for nerve versus skeletal myosin
6
a (JB032)
b (JB010)
e (JB003)
2d (JB040)
2e (JB041)
2f (JB037)
2g (JB047)
2h (JB043)
73 (1)
80 (1)
Noa
Noa
No
No
No
No
No
No
No
No
No
6
6
94 ± 7 (2)
71 (1)
1
Noa
No
1
98 (1)
1
99 (1)
No
1
94 (1)
No
1
72 (1)
Noa
Class 4: Poor potency for nerve and skeletal myosin
6f (JB004)
6g (JB012)
12b (JB039)
12i (JB036)
12a (JB038)
22 ± 10 (2)
21 (1)
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
No
No
No
No
No
No
18 (1)
16 (1)
41 (1)
1
2c (JB033)
48 (1)
a
Inhibitor concentrations > 100 µM were required to sufficiently paralyze muscle to enable action
potential assessment. N.A. not applicable.
Chemical stability, solubility and toxicity. Drug practicality is often limited by chemical and
4
9
physical properties, including toxicity, solubility, and stability. The amine side arm is important
ACS Paragon Plus Environment
2
5

Journal of Medicinal Chemistry
Page 26 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
for myosin inhibition potency and selectivity for the analogs tested in this study. Since the side
arm hydrophobicity could limit solubility, we measured kinetic solubilities of three key BHC
compounds in pH 7.4 PBS containing 2.5 % DMSO, as follows: 2, 607 ± 2; 7a, 103 ± 3; and 7i, >
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
00 M. The benzyl group in 7a has a deleterious effect on solubility, but its solubility/potency
ratio is much higher than for 1a.11, 14 As expected, the presence of the polar pyridine ring side arm
in 7i leads to higher aqueous solubility.
Most of the analogs are Schiff bases, which exist primarily as enamine tautomers in
both polar and nonpolar solvents. The imine functional group has the potential to undergo
hydrolysis under aqueous conditions. Unconjugated imines and enamines can hydrolyze under
these conditions, but BHC and its analogs are stabilized by electron delocalization. To assess
hydrolytic stability, a 100 M solution of 7i in 99:1 (v/v) PBS/methanol was monitored by UV-
Visible absorption spectroscopy during storage under ambient room conditions (light and
temperature). The half-time for hydrolysis of 7i in 99:1 (v/v) pH 7.4 PBS/methanol was found to
be 3.5 + 0.5 days (see Supporting Information). This is much longer than typical half-lives of drugs
in the body, so we conclude that the stabilities of the imine functionality in these analogs are
adequate for clinical application.
There is very little toxicity data on the BHC compounds that were previously reported in the
literature, but what is available is very encouraging. In particular, the cytotoxicities of 6b, 6j, and
7a in promyelocytic HL-60 leukemia cells have been reported as 408, 590, and 801 M,
40
respectively. This suggests that drugs developed from this structural class may have acceptable
toxicities.
ACS Paragon Plus Environment
2
6

Page 27 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
SUMMARY AND CONCLUSIONS
We have used a combination of solution ATPase and in vitro actin sliding assays, ex
vivo skeletal neuromuscular physiology, and in silico molecular docking to provide
insights into the molecular mechanisms of inhibition and binding, as well as selectivity
of BHC and a series of 34 analogs for three abundant isoforms of myosin II found in
cardiac and skeletal muscle, and in nonmuscle cells.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We provide several lines of evidence to suggest that BHC and analogs bind to the
same site as blebbistatin and have similar mechanisms of action. First, blebbistatin and
several BHC analogs inhibit the steady-state actin-activated ATPase activity of both
skeletal and cardiac myosin with similar potency and selectivity. Since this activity
represents the rate-limiting step in the ATPase cycle of myosin II, BHC analogs most
likely stabilize the actin-unbound myosin-ADP-P state, as has been previously shown for
i
blebbistatin. Second, we show that BHC, like blebbistatin, inhibits the actin sliding
velocity as measured in an in vitro motility assay. Third, an analog of BHC can compete
with blebbistatin for equilibrium binding to skeletal myosin. Finally, docking studies
with blebbistatin and BHC analogs using a template structure and two different
homology models suggest that most BHC analogs, like blebbistatin, bind within the large
cleft between the upper and lower 50 kDa sub-domains of the motor domain. Details of
the predicted binding geometries show that similar residues stabilize the binding of BHC
and blebbistatin. However, since the binding site is relatively conserved among the
ACS Paragon Plus Environment
2
7

Journal of Medicinal Chemistry
Page 28 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
myosin II isoforms studied here, modeling did not give significant insights into isoform
selectivity.
The SAR and modeling studies have identified several regions of the BHC scaffold
that tune the inhibitory potencies and therefore confer selectivity for skeletal, cardiac, or
nonmuscle myosins. A unique aspect of this study is an ex vivo skeletal muscle preparation
that allows simultaneous evaluation effects of compounds on: 1) neurotransmission, which
depends on both pre- and post-synaptic roles of nonmuscle myosin II,47, 48, 50, 51 and 2)
contractility, which depends upon muscle myosin. Combination of these data with IC50
values for ATPase inhibition of skeletal and cardiac muscle allow further SAR to emerge. A key
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
finding is that three analogs with shorter side arms (1-3 carbons) with IC ≥100 µM for
50
skeletal and cardiac myosin ATPase activity blocked neurotransmission, suggesting they
are selective for nonmuscle myosin IIB. This same pattern was also found for several
analogs with the quinolinone scaffold with shorter side arms, suggesting that nonmuscle
myosin IIB can accommodate nitrogen versus oxygen within the ring. More quantitative
information about relative potencies will be addressed in future studies with purified
nonmuscle myosin IIB.
All analogs with branched aliphatic side arms were poor inhibitors of both skeletal
and cardiac myosins, but branched side arms with more than 4 carbons were not
investigated. Adding more than 3 carbons to linear aliphatic side arms was required for
potency for both cardiac and skeletal myosins. Incorporation of an aromatic ring
ACS Paragon Plus Environment
2
8

Page 29 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
improved potency further. Replacing benzene with pyridine reduced potency, but 4-
pyridylmethyl analog 7i had the highest selectivity (≥12-fold) for inhibition of skeletal vs.
cardiac myosin. Importantly, 7i lacked potency for nonmuscle myosin IIB. Therefore, 7i
appears to be a promising candidate to relax skeletal muscle with less effect to slow the
heartbeat or to affect the many functions in nonmuscle cells mediated by nonmuscle
myosin IIB, most importantly in the peripheral and central nervous systems.52
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
While pharmacological, genetic and immunohistochemical evidence47, 50, 53 suggests that
inhibition of NMIIB, rather than NMIIA or NMIIC, likely underlies the effects of blebbistatin and
BHC analogues on neurotransmission observed in this study, we cannot rule out effects on other
non-muscle myosins such as myosin VI, which also regulates synaptic transmission and is
sensitive to blebbistatin.54
Models of BHC analogs bound to myosin at the blebbistatin site provide insights
that are generally consistent with the SAR trends summarized above. Importantly, the
side arms of BHC analogs appear to interact with hydrophobic residues lining an
extension of the binding pocket, which accommodates the longest side arm length of 6i
(6 carbons) and might accommodate longer side chains in new analogs. Future drug-
design efforts could focus on tailoring aliphatic and aromatic ring hydrophobic
interactions in the side-arm binding region to enhance potency and selectivity, and also on
adding groups of moderate polarity to adjust physical properties, such as solubility. Available
data on solubility, stability, and toxicity suggest that the BHC scaffold may be used to
develop clinical candidates.
ACS Paragon Plus Environment
2
9