Imidazodiazepine anticonvulsant, KRM-II-81, produces novel, non-diazepam-like antiseizure effects

Article abstract of DOI:10.1021/acschemneuro.0c00295

The need for improved medications for the treatment of epilepsy and chronic pain is essential. Epileptic patients typically take multiple antiseizure drugs without complete seizure freedom, and chronic pain is not fully managed with current medications. A positive allosteric modulator (PAM) of α2/3-containing GABAA receptors (5-(8-ethynyl-6-(pyridin-2-yl)-4H-benzo[f]imidazole[1,5-α][1,4]diazepin-3-yl) oxazole or KRM-II-81 (8) is a lead compound in a series of imidazodiazepines. We previously reported that KRM-II-81 produces broad-based anticonvulsant and antinociceptive efficacy in rodent models and provides a wider margin over motoric side effects than that of other GABAA receptor PAMs. The present series of experiments was designed to fill key missing gaps in prior preclinical studies assessing whether KRM-II-81 could be further differentiated from nonselective GABAA receptor PAMs using the anticonvulsant diazepam (DZP) as a comparator. In multiple chemical seizure provocation models in mice, KRM-II-81 was either equally or more efficacious than DZP. Most strikingly, KRM-II-81 but not DZP blocked the development of seizure sensitivity to the chemoconvulsants cocaine and pentylenetetrazol in seizure kindling models. These and predecessor data have placed KRM-II-81 into consideration for clinical development requiring the manufacture of kilogram amounts of good manufacturing practice material. We describe here a novel synthetic route amenable to kilogram quantity production. The new biological and chemical data provide key steps forward in the development of KRM-II-81 (8) as an improved treatment option for patients suffering from epilepsy.

Full text of DOI:10.1021/acschemneuro.0c00295

Subscriber access provided by CARLETON UNIVERSITY
Article
The Imidazodiazepine Anticonvulsant, KRM-II-81,
Produces Novel, Non-diazepam-like Antiseizure Effects
Daniel E. Knutson, Jodi L. Smith, Xingjie Ping, Xiaoming Jin, Lalit K Golani, Guanguan
Li, V. V. N. Phani Babu Tiruveedhula, Farjana Rashid, Md Yeunus Mian, Rajwana
Jahan, Dishary Sharmin, Rok Cerne, James M. Cook, and Jeffrey M. Witkin
ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.0c00295 • Publication Date (Web): 30 Jul 2020
Downloaded from pubs.acs.org on August 3, 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 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
The Imidazodiazepine Anticonvulsant, KRM-II-81, Produces Novel, Non-diazepam-like
Antiseizure Effects1
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
a
b
c
c
a
Daniel E. Knutson, Jodi L. Smith, Xingjie Ping, Xiaoming Jin, Lalit K. Golani, Guanguan
a
a
a
a
a
Li, V. V. N. Phani Babu Tiruveedhula, Farjana Rashid, Md Yeunus Mian, Rajwana Jahan,
Dishary Sharmin, Rok Cerne, James M. Cook, and Jeffrey M. Witkin a,b,e*
a
b,d
a
aDepartment of Chemistry & Biochemistry, Milwaukee Institute for Drug Discovery, University
of Wisconsin-Milwaukee, Milwaukee, WI, USA
bLaboratory of Antiepileptic Drug Discovery, Peyton Manning Hospital for Children Ascension
St. Vincent, Indianapolis, IN, USA
cDepartment of Anatomy and Cell Biology Indiana University/Purdue University, Indianapolis,
IN, USA
dInstitute of Pathophysiology, Faculty of Medicine, University of Ljubljana, Zaloška cesta 4,
Ljubljana, Slovenia.
eDepartments of Neuroscience and Trauma Research, Ascension St. Vincent Hospital,
Indianapolis, IN, USA
Footnote
1
This paper is dedicated to Dr. Phil Skolnick for his pioneering work on GABA receptor
A
pharmacology and biology, outstanding contributions to the field of neuroscience, drug
development, and the treatment of drug abuse. In gratitude for 40 years of fruitful collegial
interaction and advancements with our friend.
1
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 2 of 55
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
The need for improved medications for the treatment of epilepsy and chronic pain is essential.
Epileptic patients typically take multiple antiseizure drugs without complete seizure freedom and
chronic pain is not fully managed with current medications. A positive allosteric modulator
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
(PAM) of α2/3-containing GABA receptors (5-(8-ethynyl-6-(pyridin-2-yl)-4H-
A
benzo[f]imidazole[1,5-α][1,4]diazepin-3-yl) oxazole or KRM-II-81 (8) is a lead compound in a
series of imidazodiazepines. We previously reported that KRM-II-81 produces broad-based
anticonvulsant and antinociceptive efficacy in rodent models and provides a wider margin over
motoric side effects than that of other GABA receptor PAMs. The present series of experiments
A
was designed to fill key missing gaps in prior preclinical studies assessing whether KRM-II-81
could be further differentiated from non-selective GABA receptor PAMs using the
A
anticonvulsant diazepam (DZP) as a comparator. In multiple chemical seizure provocation
models in mice, KRM-II-81 was either equally or more efficacious than that of DZP. Most
strikingly, KRM-II-81 but not DZP blocked the development of seizure sensitivity to the
chemoconvulsants cocaine and pentylenetetrazol in seizure kindling models. These and
predecessor data have placed KRM-II-81 into consideration for clinical development requiring
the manufacture of kilogram amounts of GMP material. We describe here a novel synthetic
route amenable to kilogram quantity production. The new biological and chemical data provide
key steps forward in the development of KRM-II-81 (8) as an improved treatment option for
patients suffering from epilepsy.
Keywords: KRM-II-81, GABA receptor PAMs, 2/3-containing GABA receptors, diazepam,
A
A
epilepsy, seizure kindling
2
ACS Paragon Plus Environment

Page 3 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Introduction
Epilepsy is a chronic neurological disorder suffered by millions of people world-wide with large
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
negative health outcomes and life disruption. Many antiseizure drugs of different structural and
2
pharmacological classes have been approved for medical use. Nonetheless, seizures are often
not fully controlled despite the daily administration often of more than one antiseizure
medication. In up to 70% of epileptic patients, standard of care medicines are not efficacious.3-5
The uncontrolled seizures can also increase the probability of subsequent epileptic events
through sensitization mechanisms termed seizure kindling.6-8 Chronic, uncontrolled seizures and
the side effects arising from seizure medications have a negative impact on the developing and
adult brain and can lead to severe impairment of neurocognitive function.9, 10 Uncontrolled
seizures can also increase the liklihood of neuronal cell death and patient lethality.11 Therefore,
there continues to be an urgent need for improved antiseizure medicines.12, 13
The ligand, 5-(8-ethynyl-6-(pyridin-2-yl)-4H-benzo[f]imidazo[1,5-a][1,4]diazepin-3-yl)oxazole
or KRM-II-81 (8, Figure 1) is a positive allosteric modulator (PAM) of 2/3-
containing GABA receptors with potential for medicinal use in epilepsy based upon
A
preclinical data acquired over the past several years (Table 1). The excitement over this
molecule arises from the fact that although GABA is a long-known modulator of epilepsy,
GABA receptor PAMs like DZP are not used as daily antiseizure agents due to the
A
2
sedation and motoric side effects occurring at anticonvulsant doses. DZP
does not discriminate among GABA receptors composed of different alpha subunits, whereas
A
3
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 4 of 55
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
4, 15
KRM-II-81 is selective for 2/3-containing GABA receptors.
Genetic mutation
A
experiments combined with pharmacological studies with alpha subtype-selective molecules
have long suggested eliminating potentiation of 1-containing GABA receptors as one means
A
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
of reducing sedation and motor impairment.16 KRM-II-81 exhibits reduced sedation and motoric
burden compared to DZP.17, 18 As such there is the opportunity for achieving higher drug
exposures with KRM-II-81 to modulate GABA receptors for improved therapeutic advantage.
A
This gain in efficacy has been demonstrated with KRM-II-81 in models of pain where DZP
could not be dosed sufficiently high enough to achieve efficacy, whereas KRM-II-81 was at least
as efficacious and more potent than tramadol.19 In rodent antiseizure models, KRM-II-81 has
also shown greater efficacy than DZP under some assay conditions (Table 1). The present study
was undertaken to further differentiate DZP from KRM-II-81.
A recent licensing option agreement from RespireRx to acquire this molecule from the
University of Wisconsin-Milwaukee established increased prioritization toward the development
of KRM-II-81 for epileptic patients. Although the case for development of KRM-II-81 as an
antiseizure agent is gaining traction (Table 1), there are some gaps in the biology. As an
anticonvulsant against chemically induced seizures, KRM-II-81 has only been studied against
pentylenetetrazol (PTZ). However, since PTZ is a GABA receptor antagonist, the blockade of
A
seizures with a GABA receptor PAM like KRM-II-81 or DZP is expected on pharmacological
A
grounds; blockade of seizures might be due to receptor pharmacology rather than to suppression
of the seizurogenic effects of PTZ per se. The present study addresses this deficiency by
studying a broader array of chemoconvulsants with diverse non-GABA receptor mechanisms,
A
4
ACS Paragon Plus Environment

Page 5 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
thus enabling a pure test of the antiseizure hypothesis and side-by-side comparisons with
diazepam.
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
KRM-II-81 (8) has also shown an ability to block seizures in rodent models where seizures have
been developed over time by sensitizing seizure networks by daily brain stimulation, a process
known as seizure kindling. Seizure kindling can be modeled in rodents and is often used as
another means of differentiating antiseizure drugs.20, 21 The data in Table 1 summarize the
ability of KRM-II-81 to suppress the expression of seizures that have already been developed
(blockade of seizure expression). What remains unknown is whether KRM-II-81 can block the
development of the seizure sensitization process (kindling). In contrast to other GABA receptor
A
PAMs like some neuroactive steroids, DZP has limited impact on this process.22, 23 In order to
determine the relative efficacy of KRM-II-81 on seizure kindling, the effects of DZP and KRM-
II-81 were compared in two kindling models in the present study.
Another obstacle to the development of KRM-II-81 (8) for epileptic patients is the synthesis of
GMP material on large scale for IND-enabling toxicology and for first human dose studies.
Previous work has created synthetic routes from the precursor molecule HZ-166 (6, Scheme 1).
The first route employed a three step approach which involved a low yielding LAH reduction of
the ester (HZ-166) to an alcohol which was then subsequently oxidized by manganese dioxide to
the aldehyde 7.15 A modification of this synthesis improved these final steps of the synthesis
by converting HZ-166 (6) to its respective ester bioisostere KRM-II-81 (8) in two steps by
switching the reduction reagent to PDBBA to afford the key aldehyde 7 directly. However, a
process conducive to larger scale chemistry was still required to synthesize the ester precursor
24
5
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 6 of 55
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
(
HZ-166) to KRM-II-81. We sought to replace the chromatographic purification steps employed
in all five steps of the earlier discovery process for the synthesis of HZ-166 with crystallizations.
In addition, to address scale up issues including poor stirring, side product formation, long
process times and yield losses on large scale. Provided herein is a method amenable for
production of kilogram quantities of HZ-166 (6) and KRM-II-81 (8).
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
ACS Paragon Plus Environment

Page 7 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Results and Discussion
KRM-II-81 has Broader Efficacy Against Chemical Convulsant Agents than DZP. In order
to extend the observations on the anticonvulsant activity of KRM-II-81 (8) to other, non-GABA
convulsants, chemicals that act upon monoamine transporters (cocaine), potassium channels (4-
aminopyridine or 4-AP), N-methyl-D-aspartate (NMDA) receptors (NMDA), glycine receptors
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
(strychnine), and muscarinic cholinergic receptors (pilocarpine)(acute seizures, not status
epilepticus) were studied. In addition, PTZ and another antagonist of GABA receptors,
A
picrotoxin, were also studied. The comparative efficacy of DZP (1 mg/kg, i.p.) and KRM-II-81
(30 mg/kg, i.p.) are summarized in Error! Reference source not found.. The doses of DZP and
KRM-II-81 were based upon prior data showing that these doses represent ~ED values for seizure
9
5
blockade against PTZ-induced clonus in mice.18, 22 The doses of both DZP and KRM-II-81 are
further justified for this single dose comparison study based upon the equivalent efficacy of these
two anticonvulsants against PTZ-induced clonus, tonus, and lethality (Table 2).
For PTZ, both DZP and KRM-II-81 significantly attenuated clonic convulsions, as previously
reported in rats.18 Prior studies with PTZ did not study doses inducing tonic seizures and
lethality. Here we show that both DZP and KRM-II-81 (8) also significantly reduce these toxic
endpoints induced by PTZ. Although a similar profile of comparative protection occurred when
studying picrotoxin, the statistical preference for KRM-II-81 might be due to the relatively small
percentage of tonic convulsions and lethality that were produced by the dose of picrotoxin
18
employed. As with the GABA receptor antagonists PTZ and picrotoxin, both DZP and KRM-II-
A
8
1 were also equally effective against the acute seizures evoked by the muscarinic cholinergic
receptor agonist pilocarpine.
7
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 8 of 55
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
In contrast, for cocaine (COC), 4-AP, NMDA, and strychnine, KRM-II-81 was efficacious when
DZP was not. Clonic convulsions were blocked by KRM-II-81 for these convulsant compounds.
The differential efficacy of KRM-II-81 over DZP extended also to tonic seizures and lethality in
the case of 4-AP, strychnine, and NMDA. The findings reported here are consistent with the
literature on the anticonvulsant effects of DZP.25-30 Overall, the data support the conclusion that
KRM-II-81 has broad ranging anticonvulsant efficacy against chemoconvulsants (Error!
Reference source not found.). Further, the data in Error! Reference source not found. support
the conclusion that KRM-II-81, in the present study, has a superior efficacy profile over that of
DZP as shown earlier with other seizure provoking stimuli (Table 1). Statistically significant
blockade (Fisher’s Exact probability test) were observed for KRM-II-81 but not DZP against COC
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
(clonus), 4-AP (clonus and tonus), NMDA (clonus and lethality), picrotoxin (tonus and lethality),
and strychnine (clonus, tonus, and lethality). This differentiation is most striking for 4-AP, NMDA,
and strychnine especially since doses of DZP and KRM-II-81 produced equivalent anticonvulsant
effects against the GABA-based chemoconvulsants PTZ and picrotoxin (Table 2). However, full
dose-response comparisons (both of chemoconvulsant and of anticonvulsant) will be ideally
required to fully appreciate precise quantitative differences between DZP and KRM-II-81 as has
been reported earlier in some anticonvulsant-detecting assays.18
KRM-II-81 but Not DZP Blocks the Development of PTZ Kindling. When PTZ, in a
subconvulsant dose (45 mg/kg), is given every other day for 4 days, the percentage of mice
2
2
th
exhibiting clonic seizures increases (Figure 2A) as previously reported. On the 5 experimental
session, either vehicle, DZP (1 mg/kg, i.p.), or KRM-II-81 (30 mg/kg, i.p.) was given prior to PTZ.
8
ACS Paragon Plus Environment

Page 9 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Both DZP and KRM-II-81 prevented clonic seizures in these PTZ-sensitized mice (Figure 2A)
thus demonstrating efficacy to block seizures in fully kindled animals. In order to test whether
these compounds also prevent the expression of seizure kindling, either vehicle, DZP or KRM-II-
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
8
1 was administered prior to each dose of PTZ. Under these conditions, both drugs fully protected
mice from the occurrence of seizures and the increased sensitivity that develops to PTZ in the
absence of treatment (Figure 2B). Another experiment was conducted in order to assess whether
these compounds can prevent or attenuate the development of kindling. In this study, vehicle,
DZP, or KRM-II-81 was given on each of the first 4 experimental sessions prior to PTZ. Then, on
day 10 (experimental session 5), mice were given PTZ alone (with vehicle pretreatments). While
prior exposure to KRM-II-81 + PTZ inoculated mice from kindling, DZP did not (Figure 2C).
Thus, despite the ability of DZP to block both acute and sensitized effects of PTZ, DZP was not
able to dampen the seizure kindling process where KRM-II-81 was protective.
Another study was conducted in order to extend these observations to another seizure kindling
agent. Cocaine (COC) was chosen since the DZP and KRM showed differential efficacy against
PTZ and COC convulsant challenge. Whereas in the case of PTZ, DZP and KRM-II-81 were
equally effective in preventing seizures induced by acutely administered PTZ, only KRM-II-81
was efficacious against acutely-administered COC (Error! Reference source not found.).
Seizure kindling with COC was studied as with PTZ kindling with the exception that COC kindling
was conducted every day31, 32 vs. the every other day dosing method that is established for PTZ
kindling.22 In this study, COC given daily for 5 days produced significant and large changes in
the percentage of mice exhibiting clonic convulsions on day 5 vs. day 1 (Figure 3A) demonstrating
seizure kindling. On day 6, either vehicle, DZP (1 mg/kg, i.p.) or KRM-II-81 (30 mg/kg, i.p.) was
9
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 10 of 55
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
given prior to COC challenge. KRM-II-81 but not DZP significantly attenuated clonic seizure
incidence in mice (Figure 3B) as was observed after acute COC challenge (Error! Reference
source not found.). Thus, both in non-kindled mice and fully kindled mice, KRM-II-81 but not
DZP is an effective seizure protectant. The expression of kindled seizures over the 6-day period
of COC dosing was assessed next. KRM-II-81 significantly attenuated the expression of COC
seizures across 6 days of COC dosing; DZP was not active (Figure 3B).
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
A final seizure kindling study was initiated to determine whether DZP or KRM-II-81 (8) could
prevent the development of COC seizure kindling. In this study, vehicle, DZP, or KRM-II-81 was
given prior to each COC kindling session on days 1-5. On day 6, COC alone was given (+ vehicle)
to ascertain whether kindling was suppressed by the prior drug treatments. On the day 6 test,
KRM-II-81 but not DZP significantly attenuated the development of kindling (Figure 3C). The
blockade of kindling by KRM-II-81 appeared to be greater for PTZ kindling (Figure 2C) than that
observed with COC kindling (Figure 3C) perhaps due to the more prominent effects of KRM-II-
8
1 on acute PTZ vs. COC seizures (Error! Reference source not found.) and the greater ability
to block expression of PTZ vs. COC kindling (Figure 2B vs. Figure 3B). However, it is critical
to emphasize that even though DZP is an effective anticonvulsant against acute seizures induced
by PTZ (Error! Reference source not found.) and is efficacious as a suppressor of the expression
of PTZ kindling, DZP is still not effective as a blocker of the development of kindling while KRM-
II-81 is. The data for DZP reported here are consistent with prior findings.22, 23, 31, 32 Overall, the
kindling experiments document the differential ability of KRM-II-81 to significantly suppress the
development of the seizure sensitization process and that this effect is not confined to the GABAA
receptor antagonist convulsant PTZ.
10
ACS Paragon Plus Environment

Page 11 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
A Synthetic Method Amenable for the Production of Kilogram Quantities of KRM-II-81.
Given the potential therapeutic value of KRM-II-81 for patients as discussed above and
elsewhere,17, 18, 33 a novel method for the synthesis of KRM-II-81 with efficiency on a large scale
was developed. The synthesis of KRM-II-81 (Scheme 1) has been reported previously15, 24, but
the discovery chemistry was not well-suited for scale up in a potential GMP manufacturing setting.
All five steps of the synthesis to the key intermediate HZ-16634, 35 required purification by column
chromatography, while several steps exhibited typical scaleup issues such as inefficient stirring,
side product formation, prolonged process times and yield loss upon scale up.
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 challenge in the first step of the synthesis was to improve upon the stirring difficulties of the
reaction media and the lack of an adequate purification procedure for multigram quantities. The
36
37
discovery route developed by Li et al. based on the work of Selnick et al. called for cooling the
o
mixture of 2'-pyridyl ketone 1, sodium bicarbonate and dichloromethane to 0 C prior to the
addition of bromoacetyl bromide. Early into the addition process an extremely thick red
gelatinous-like solid developed that made efficient stirring of the reaction mixture nearly
impossible. Upon warming to room temperature, the gelatinous solid would slowly dissipate,
which resulted in a more readily stirred reaction mixture. Any procedure that is being developed
for large-scale synthesis cannot involve agitation issues. Similar research in our chemistry group
utilized the same reaction protocol,34, 38 but those analogs lacked the 2'-pyridine function and
therefore lacked the stirring difficulties encountered herein. Upon consideration of a possible
reaction mechanism (Scheme 2), it was proposed that formation of the desired 2-halooacetamide
product 2c must proceed through the insoluble acetyl pyridinium intermediate 2b for which the
11
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 12 of 55
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
red, gelatinous-like solid can be attributed. As a remedy, the initial reaction temperature was raised
o
to 25 - 30 C prior to the addition of bromoacetyl bromide, while the bromo acetyl reagent was
added over a 60minute time period. Utilization this new approach completely avoided the agitation
issue while the red, insoluble pyridinium intermediate 2b was not observed to a significant amount.
Presumably, the elevated process temperature increased the reaction rate enough, so the reaction
did not stall at the pyridinium salt stage. Analysis of this reaction progress immediately after the
addition of the bromide by TLC confirmed this observation for there was no evidence of the 2'-
pyridyl ketone 1 nor pyridinium intermediate 2b. Upon completion of the reaction, the inorganic
salts were partitioned into the aqueous layer, after which the solvent was exchanged to ethanol to
precipitate the pure halooacetamide product 2a or 2c (individually). No further purification was
required. The employment of 2-chloroacetyl chloride as an economical substitute for 2-
bromoacetyl bromide proved to perform equally as well in the reaction protocol with no
detrimental effects to the chemistry in the following steps. In summary, the modifications
employed to the synthesis provided 125.6g of 2c in 98.5% yield (2L reaction flask).
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 second step of the synthesis furnished the diazepine core 3 via the conversion of the
chloroacetamide 2c into a β-amino amide intermediate 3b prior to condensation of the amine with
the carbonyl function. The process problems presented in this reaction sequence included the lack
of a reliable and efficient method for purification of the amide 3 and the formation of detrimental
side products observed on scale up. The discovery route (Scheme 3) utilized saturated methanolic
ammonia to furnish the β-amino amide intermediate 3b in situ by nucleophilic aliphatic
substitution of the alkyl halide 2a which then should cyclize via a carbinolamine condensation to
yield the 2'-pyridyldiazepine 3. As is typical with scale-up chemistry, process times (heating and
12
ACS Paragon Plus Environment

Page 13 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
cooling) and reaction times are prolonged, as compared to smaller scale procedures and a
subsequent drop in yield of the amide 3 was observed. An investigation of the causes of the yield
reduction revealed the formation of a significant side product that was observed to a small extent
by TLC on small scale but increased substantially as the scale was increased. The isolation of the
impurity (m/z = 314.0167) indicated a mass unit one less than that of the desired product 3 (m/z =
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
13
3
15.0007). A thorough analysis of the structure of the impurity using H NMR, C NMR, HSQC,
1
13
1
15
H- C-HMBC, H- N-HMBC spectroscopy experiments provided evidence which supported the
1
15
proposed imidate 3c as the structure of the impurity. The H- N-HMBC experiment was
especially revealing because examination of the spectrum indicated the identity of the impurity 3c
contained 4 nitrogen atoms in the structure (Figure S1). The pyridine nitrogen atom (N13)
correlated to pyridine ring protons (H11 and H12), while the remaining nitrogen atoms (N15, N18
and N19) correlated to the aliphatic CH protons (H16). Finally, the iminium nitrogen atom (N18)
2
correlated to the A-ring proton, H6. The logical explanation for the formation of the imidate (3c)
would be through a dehydration mechanism first involving addition of ammonia into the diazepine
amide carbonyl via nucleophilic acyl substitution, which was followed by the elimination of water
(Scheme S1). The mechanism seems reasonable because of the extended reaction times under
high concentrations of ammonia. Options to mitigate the formation of the imidate 3c included
shorter reaction times and lower ammonia concentrations, but the application of these solutions
were ruled out due to concern for low yields attributed to incomplete consumption of the starting
material, which would then have to be removed.
Clearly, a new approach to synthesis of the key 2'-pyridyl-diazepine core (3) was warranted. Here,
one turned to the Delépine reaction developed by Stéphane Marcel Delépine.39 It entails
13
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 14 of 55
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
incorporation of the hexamethylenetetramine-(HMTM)-based cyclization reaction first developed
4
0
41
by Blažević and Kajfež and further refined by Cepanec, et al. The HMTM-based cyclization
(Scheme 4) involves first, the formation of quaternary ammonium salt 3a by nucleophilic aliphatic
substitution of the alkyl halide 2 and this was followed by in situ hydrolysis to the β-amino amide
intermediate 3b, which was then condensed with the carbonyl function to provide the 2'-pyridyl-
diazepine 3. This new approach performed well albeit with slightly lower than expected yields
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
(
60-65%) than the literature report.41 An investigation into the side products revealed a 5-10 %
impurity identified as β-alkoxy amide 3d. Presumably, when a primary alcohol (methanol or
ethanol) was used as the solvent this undesired Sn2 addition of the alcohol to displace the tetramine
to provide to 3d was possible. Modification of the reaction solvent to the secondary alcohol, 2-
propanol (IPA) eliminated this side reaction evidenced by the lack of detection of the
corresponding β-alkoxy amide impurity 3d; importantly, the corresponding yield increase was
realized. On large-scale (2L flask) the utilization of the Delépine reaction delivered 75g (75%
yield) of 2'-pyridyl-benzodiazepine 3.
The third step of the synthesis of HZ-166 (6) was to install the imidazole ring C to produce
4
2
imidazodiazepine 4. The previous procedure involved formation of the iminophosphate 4a
(
Scheme 5) by deprotonation of the starting amide using potassium t-butoxide followed by the
addition of diethylchlorophosphate. Then, iminophosphate 4a subsequently underwent a
nucleophilic addition with the enolate of ethyl isocyanoacetate to furnish the desired
imidazodiazepine 4. Several improvements to the original chemistry were required in order to
provide a reliable and scalable procedure. Initially, the purification was conducted using column
chromatography. Secondly, the reagents were added in an “all in one” fashion and specifically as
14
ACS Paragon Plus Environment

Page 15 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
a solid in the case of the portions of potassium t-butoxide. Thirdly, each subsequent reagent
o
addition involved cooling of the reaction mixture to below -35 C. Then, the next reagent was
o
added and finally the mixture was allowed to warm to 0 C for each constituent of the process to
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
react. With those challenges in mind we sought to develop a purification method by crystallization
and to determine a reaction temperature in which the reagents could be added in a controlled
fashion at a temperature at which they would react. By careful monitoring of the reaction progress
by TLC it was determined that the iminophosphate 4a did not form until the reaction temperature
o
reached -20 C. Similarly, it was determined that the required reaction temperature of the
o
condensation of the enolate of ethyl isocyanoacetate with the iminophosphate 4a was -35 to -30 C.
Consequently, it was felt that addition of the reagents in a controlled, dropwise fashion would
o
allow one to simply hold the reaction mixture at -20 to -15 C throughout the sequence of additions.
This procedure first involved dissolving diazepine 3 in tetrahydrofuran (THF) and cooling that
o
solution to -20 C. Then, a solution of potassium t-butoxide (1.3 eq) in THF was added dropwise
over 30 minutes to the reaction mixture. The reaction mixture was then stirred for 30 minutes at -
o
2
0 to -15 C to ensure complete deprotonation of the amide starting material. Next,
diethylchlorophosphate (1.4 eq) was added over a 15-minute time span. Within 2 hours after the
addition of the phosphate reagent, consumption of the amide starting material 3 was observed to
generate the iminophosphate intermediate 4a. Then, ethyl isocyanoacetate (1.3 eq) was added over
1
5 minutes, and this was followed by a dropwise addition of a solution of potassium t-butoxide
o
(1.3 eq) in THF at -20 to -15 C. The isocyanoacetate / butoxide addition sequence was designed
in this manner so that the enolate would react with the iminophosphate as readily as it formed. As
predicted, the formation of imidazodiazepine 4 was complete immediately after the t-butoxide
addition with no detectable iminophosphate intermediate 4a observed. Upon completion of the
15
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 16 of 55
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
reaction procedure the imidazodiazepine product 4 was partitioned into the organic layer using
dichloromethane. An aqueous bicarbonate work-up removed the potassium salts and unreacted
butoxide. The purification of the product was achieved by crystallization from t-butyl methyl ether
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
(tBME), which adequately removed any traces of side products related to the phosphate related
byproducts. The trituration of the crystals with hot ethanol was then employed to remove any
residual diazepine starting material 3, if necessary. With the goal of eliminating the column
chromatography and a more robust and scalable procedure in hand, successful execution of the
large-scale synthesis provided 61.0 g (52 % yield) of imidazodiazepine 4 (5 L flask).
The next step in the reaction sequence employed the copper-free Sonogashira coupling43, 44 to
covert aryl bromide 4 into the silyl-protected acetylene 5. The earlier discovery chemistry used
TMS-acetylene (1.5 eq) and bis(triphenylphosphine)palladium(II) diacetate (Pd(OAc) (PPh ) ) as
2
3 2
the catalyst, as well as a large excess of triethylamine. Under these conditions the reaction time
o
was typically 15 hours at reflux (75 C) to achieve the desired consumption of the aryl bromide 4
(Table 3, entry 1). These reaction conditions presented an opportunity for improvement. First,
the catalyst was prone to poisoning and the procedure required an extensive de-gassing protocol
prior to initiating the reaction. The degree of proper degassing (to remove dissolved oxygen in the
solvents) played a direct role on the success or failure of the coupling reaction. Secondly, the
reaction conditions employed an excess of TMS-acetylene (1.5 eq) to drive the reaction to
completion, while most often catalytic poisoning and perhaps degradation of the TMS-reagent
stalled the reaction with 5-10% of aryl-bromide 4, which remained. Typically, additional TMS-
acetylene and fresh catalyst were added to complete the reaction. Thirdly, solvent quantities of
triethylamine (coupled with acetonitrile in a 1:1 mixture) were used to dissolve the
16
ACS Paragon Plus Environment

Page 17 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
imidazodiazepine 4 starting material in the reaction medium.
Finally, difficult column
chromatography was required to purify the product to remove the closely eluting aryl bromide 4
starting material from the crude mixture. This process was also further complicated by the
triethylammonium bromide side product, which remained in the crude residue by the previous
processing protocol. Therefore, in order to develop a reliable large-scale procedure for the copper-
free Sonogashira coupling, the catalyst poisoning, the acetylene reagent, the isolation process and
the purification of the silyl target 5 must be improved.
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
To remedy the problem of the poisoning of the catalyst, which was caused by the oxidation of
triphenylphosphine to the corresponding phosphine oxide, a ligand less prone to oxidation was
required. A trial with tri-ortho-tolylphosphine (P-o-tol ) developed by Greg Fu was first attempted
3
o
o 45, 46
because it has a larger cone angle (194 ) than that of triphenylphosphine (PPh , 145 )
and
3
should exhibit improved resistance to oxide formation and subsequent catalyst poisoning. As
mentioned, the ligand in the process was altered to P-o-tol , which significantly shortened the
3
reaction time to 6 hours (from 15 hours) and no cumbersome degassing was required (Table 3,
entry 2). The starting TMS-acetylene (1.5 eq) could be replaced with the inherently more stable
TIPS-acetylene reagent (1.2 eq), which shortened the reaction time further to 4 hours, even with
reduced equivalents of the protected-acetylene reagent. (Table 3, entry 3). Finally, reduction of
triethylamine from the solvent quantities to stoichiometric quantities (2.0 eq) exhibited no
detrimental impact on the reaction time (Table 3, entry 4). With the optimal reaction conditions
in hand the product isolation and purification procedure were then refined.
The spent and
undissolved catalyst was filtered from the reaction media through a small pad of silica gel. The
filtrate, which resulted, was concentrated and then partitioned between dichloromethane and
17
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 18 of 55
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
aqueous sodium bicarbonate to remove triethylammonium bromide. Finally, the concentrated
residue was run through a short flash column to remove the baseline impurities. No other
purification was needed in order to carry the product 5b forward into the next step since the aryl
bromide starting material 4 was completely consumed using the refined reaction conditions. In a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
L flask, the improved copper-free Sonogashira coupling process produced 65.0g of the TIPS-
protected acetylene 5b in 85% yield.
The final step of this development process provided the key intermediate HZ-166 (6) via
deprotection of the TIPS-acetylene function using fluoride anion. The improvements to this
process included optimization of the reaction temperature and replacement of the chromatographic
purification with crystallization. The discovery process utilized tetrabutylammonium fluoride
(TBAF) as the fluoride source. This reagent performed, as required, but the isolation and
purification of HZ-166 target (6) required removal of the tetrabutylammonium hydroxide
byproduct, adequately. Additionally, the purification protocol would also need to efficiently
o
remove the TIPS-fluoride side product. The discovery route cooled the reaction media to -78 C
prior to the addition of TBAF. Careful monitoring of the reaction by TLC indicated that
o
deprotection did not initiate, however, until the temperature reached -20 C. Consequently, the
o
process temperature was modified to – 20 to -15 C, at which point the TBAF·xH O (1.0M in THF)
2
was added over a 30-minute time period. The process was monitored by TLC, which indicated
complete deprotection had occurred upon completion of the TBAF addition.
Evaluation of a variety of solvents led to the selection of 2-propanol (IPA) as the ideal
crystallization solvent, which exhibited minimal product solubility with efficient side-product
18
ACS Paragon Plus Environment

Page 19 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
(
TBA-OH and TIPS-F) solubility. The improved final step of the HZ-166 (6) synthesis employed
the optimum temperature, while eliminating the chromatographic purification providing the 38.4g
(85.0%) of HZ-166 (3L flask).
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
In summary, the five-step synthesis of HZ-166 (6) was refined to provide a more reliable and
robust procedure (Scheme 6) with a 27.6% overall yield. As noted, the chromatographic
purification steps were replaced with large-scale crystallizations in four steps, while importantly,
the overall reaction times were reduced. The conversion of HZ-166 (6) on 20 gram scale into
KRM-II-81 (8) involved first the reduction of the ester with the hindered reducing agent potassium
o
diisobutyl-tert-butoxyaluminum hydride (PDBBA) at 0 C to provide the aldehyde (7) in 80%
yield. This was followed by the reaction of the aldehyde (7) with toluenesulfonylmethyl
isocyanide (TosMIC) to afford the oxazole, KRM-II-81 (8) in 89% yield. This optimized
2
4
conversion was previously carried out by Li, et al, and the experimental of which is fully detailed
in the Supporting Information.
As already discussed, the need for improved antiseizure drugs is critical to the medical and life
needs of epileptic patients. The ability of KRM-II-81 (8) to produce anticonvulsant effects in
multiple preclinical models and to generate, in some cases, superior efficacy to that of another
GABA receptor PAM, DZP, supports the proposition that KRM-II-81 represents a new lead
A
candidate for development. KRM-II-81 was also recently shown to block the cortical hyperactivity
1
7
of neurons after traumatic brain injury in a mouse model, which suggests potential efficacy
against the development of post-traumatic epilepsy.47, 48 The data in models of pharmacoresistant
epilepsy (Table 1) combined with the current findings in kindling experiments (Figure 2C and
Figure 3C) provide important new data on compound superiority. Ganaxolone is a neuroactive
19
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 20 of 55
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
steroid that has shown similar antiepileptogenic activity against COC and PTZ kindling under
conditions where DZP is not effective.22, 23, 32 Ganaxolone was recently assessed in patients with
difficult to treat status epilepticus, a life-threating condition. In this study, ganaxolone was
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
efficacious, thus translating preclinical predictions into patients in this emergency situation.49
In addition to the data in rodent seizure models presented here and elsewhere,17, 18 additional
patient translational assurance comes from the efficacy of KRM-II-81 (8) to dampen
hyperexcitabity of neural networks in cortical tissue slices from epileptic patients (Table 1).
PF‐06372865 is another compound that has selectivity for α2/3-containing GABA receptor
A
5
0
although, unlike KRM-II-81, PF-96372865 also potentiates 5-containing GABA receptors.
A
PF‐06372865 is well-tolerated in people50-52 and was efficacious in inhibiting electrical activity in
patients with photosensitive epilepsy.51
The more benign sedative and motor-impacting effects of KRM-II-8115, 18, 33 are an additional
feature of its pharmacology that bode well for clinical superiority. Recent docking studies of
KRM-II-81 (8) with the CryoER structure 6HUP (α1β3γ2L GABA receptor in complex with
A
53
DZP) provided structural support for the reduced impact of KRM-II-81 at the α1His102 side
chain implicated in sedation and motor-impairment.17 KRM-II-81 also showed less respiratory
depression than alprazolam.18 Additional data suggest that 2/3-selective PAMs will have less
liability than non-selective GABA receptor PAMs to produce memory impairment, tolerance, or
A
abuse.54-58 KRM-II-81 is also predicted to have efficacy in other therapeutic domains that are both
stand-alone diseases and are comorbid with epilepsy including anxiety,15, 59 depression, and
60
pain.14, 19
20
ACS Paragon Plus Environment

Page 21 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
In conclusion, the data presented in the present manuscript along with literature reports are
consistent with KRM-II-81 (8) showing superiority over diazepam and being a lead compound for
clinical development for pharmacoresistant epilepsy and other epileptic states. The newly
designed synthesis amenable to kilogram quantities of material presented here provides another
key step in the drug development process.
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
21
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 22 of 55
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
Experimental Section
Biology
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
Compounds. KRM-II-81 (8) was synthesized (the laboratory of James M. Cook) as previously
described.15, 24 The other compounds were obtained from Sigma-Aldrich (St. Louis, MO, USA).
KRM-II-81 was suspended in 1% carboxymethylcellulose and dosed at 1 ml/kg in rats below doses
of 30 mg/kg; 30 mg/kg (dosed at 3 ml/kg), 60 mg/kg (dosed at 6 ml/kg). Mice were dosed in a
volume of 10 ml/kg. The other compounds were dissolved in sterile water with sonication as
needed. Compounds were dosed i.p. with the exception of pentylenetetrazol (PTZ), which was
given by s.c. injection in the acute convulsant tests.
Rodent Assays. All studies were performed in accordance with the guidelines of the National
Institutes of Health and by local animal care and use committees. The local animal care and use
committee and veterinary staff provided direct oversight of the animals by inspections, protocol
reviews, laboratory site visits, and animal health monitoring. Male, CD1 mice were used and
weighed 28-33g at the time of testing. The mice were allowed to acclimate to the vivarium for at
least 4 days prior to testing and for at least 45 min in the testing room prior to experimentation.
Animals were housed in a temperature- and humidity-controlled room with a 12 h light/dark cycle
(on at 0600). Mice were group housed in large plastic containers with sawdust bedding. Standard
mouse food pellets and water were continuously available except during experimental sessions.
Mice used in the acute convulsant studies were used only once and then euthanized. Mice in the
kindling studies were used only for their respective test group and then euthanized after the last
test session.
22
ACS Paragon Plus Environment

Page 23 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Chemoconvulsant Studies. The experiments were conducted with doses and routes of
administration per the general protocol of Witkin et al. 61 with the exception that PTZ was given
at 75 mg/kg by the s.c. route and that the mice were a different outbred strain than previously
employed. Mice were given either vehicle, diazepam (DZP) (1 mg/kg, i.p.) or KRM-II-81 (30
mg/kg, i.p.) and placed into a holding cage for 30 min. DZP and KRM-II-81 studies for each
chemoconvulsant were run in side-by-side experiments on the same day. The doses of DZP and
KRM-II-81 have been shown previously to produce nearly full protection against PTZ-induced
clonic convulsions.18, 22, 23 After 30 min, the mice were given the convulsant agent and placed into
individual observation chambers where they were observed for 60 min for clonic convulsions,
tonic convulsions and lethality by trained observers. The percentage of mice exhibiting clonus,
tonus, and lethality were recorded. Statistically significant differences from vehicle treatments
were assessed by Fisher’s Exact Probability test with p<0.05 as an a priori level of significance.
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
23
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 24 of 55
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
Seizure Kindling. Seizure kindling was established in separate groups of mice by giving close to
subconvulsant doses of cocaine (COC) (60 mg/kg, i.p.) or PTZ (45 mg/kg), as previously described
for COC31, 32 where cocaine was given every day and for PTZ,22, 23 where PTZ was given every
other day. Each experiment was independently conducted twice with groups of 6 mice each. For
PTZ seizure kindling, PTZ was given on days 1,3,5, and 8 and then the mice were tested with PTZ
again on day 10. For COC, COC was given on days 1,2,3,4, and 5 and then tested with COC on
day 6. Three separate groups of mice were studied in each of three separate experiments to study
drug effects on 1) fully kindled seizures, 2) the expression of kindled seizures, and on 3) the
development of kindling where either vehicle, DZP (1 mg/kg, i.p.), or KRM-II-81 (30 mg/kg, i.p.)
was studied. Fully-Kindled Seizures: vehicle + PTZ or COC were given on kindling days/vehicle,
DZP, or KRM-II-81with PTZ or COC given on test day; Expression of Kindling: Vehicle + PTZ
or COC, DZP + PTZ or COC, or KRM-II-81 + PTZ or COC was given for each kindling day and
for the test day; Development of Kindling: Vehicle, DZP, or KRM-II-81 + PTZ or COC was given
on each kindling day/Vehicle + PTZ or COC was given on the test day. Data presented are the
percentage of mice exhibiting convulsions; mice were not scored for seizure severity. The data
were analyzed by two-way ANOVA analyzing treatment and treatment day with repeated
measures on subjects. Bonferonni post-hoc tests compared each treatment day to its vehicle
control (p<0.05 considered to be statistically significant a priori).
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
24
ACS Paragon Plus Environment

Page 25 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Chemistry
General Procedures for the Synthesis of HZ-166 (6). All reactions were performed in round-
bottom flasks with magnetic stir bars or overhead mechanical stirrers under an argon atmosphere.
Organic solvents were purified when necessary by standard methods or purchased from Sigma-
Aldrich Chemicals were purchased from either Sigma Aldrich, Oakwood Chemical, Alfa Aesar,
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
13
Matrix Scientific, Admiral Chemical Company, or Acros Organic. The H and C NMR data
were obtained on Bruker Spectrospin 300 MHz and 500 MHz instruments with the chemical shifts
in δ (ppm). The HRMS spectral data was obtained on a LCMS-IT-TOF by Shimadzu Scientific.
25
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 26 of 55
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
N-(4-Bromo-2-picolinoylphenyl)-2-chloroacetamide (2c). To a mixture of (2-amino-5-
bromophenyl)(pyridine-2yl)methanone (1, 100 g, 360.9 mmol), sodium bicarbonate (60.6 g, 721.7
mmol), and dichloromethane (1000 mL), chloroacetyl chloride (43.0 mL, 541.3 mmol) was added
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
o
dropwise over 60 min while keeping the temperature between 25 – 30 C. The bright red reaction
mixture, which resulted, was then allowed to stir for more than 1 h at rt. The completion of the
reaction was verified by analysis by TLC using silica gel and 50% ethyl acetate / hexanes. The
reaction mixture was then slowly diluted over 30 min with water (500 mL) as carbon dioxide gas
evolved. The biphasic mixture, which resulted, was allowed to stand for 15 min and the layers
were separated. The aq layer was extracted with dichloromethane (500 mL) and the combined
organic layers were washed with 5% aq sodium bicarbonate solution (500 mL) and then 10% aq
sodium chloride solution (500 mL). The organic layer was dried (Na SO ). The solvents were
2
4
o
removed under reduced pressure and the residue was slurried with ethanol (500 mL) at 50 – 55 C
for 30 min. Upon cooling to rt and after holding the temperature for 1 h, the solid was filtered and
o
washed with ethanol (100 mL x 3). The solid was dried under vacuum at 40 C to afford the product
nd
2
c as an off-white solid (115.2 g, 90.3%). Additional product (2c) was obtained as a 2 crop by
concentrating the filtrate (10.4 g, 8.2%). Total yield: 125.6g, 98.5%: Rf = 0.6 (EtOAc-hexanes,
1
1
1
7
:1 and 1% of TEA); H NMR (500 MHz, CDCl ) δ 11.66 (s, 1H), 8.76 (ddd, J = 4.8, 1.7, 0.9 Hz,
3
H), 8.58 (d, J = 9.0 Hz, 1H), 8.03 (dt, J = 7.8, 1.1 Hz, 1H), 8.01 (d, J = 2.4 Hz, 1H), 7.96 (td, J =
1
3
.7, 1.7 Hz, 1H), 7.72 (dd, J = 9.0, 2.4 Hz, 1H), 4.21 (s, 2H); C NMR (126 MHz, CDCl ) δ
3
195.29, 165.34, 154.76, 148.83, 138.80, 137.43, 137.19, 136.88, 126.70, 124.99, 124.62, 122.99,
1
3
15.81, 43.15; HRMS (ESI/IT-TOF): m/z [M + H]+ calcd for C H BrClN O : 352.9693; found:
14 11 2 2
52.9690.
26
ACS Paragon Plus Environment

Page 27 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
7
-Bromo-5-(pyridin-2-yl)-1,3-dihydro-2H-benzo[e][1,4]diazepin-2-one (3). A mixture of N-
(4-bromo-2-picolinoylphenyl)-2-chloroacetamide
(2c,
115
g,
325.2
mmol),
hexamethylenetetramine (HMTM, 100.3 g, 715.5 mmol), ammonium acetate (55.2 g, 715.5
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
o
mmol), and isopropanol (2000 mL) was heated to reflux (82 C). The reaction mixture was held at
reflux for 4 h at which point the reaction was deemed complete by analysis by TLC (silica gel and
o
5
0% ethyl acetate / hexanes). The reaction mixture was then cooled to 0 – 5 C using an ice bath.
The solid, which resulted, was filtered and washed with cold isopropanol (100 mL x 2) and then
o
water (100mL x 4). The solid was dried under vacuum at 40 C to afford the product 3 as an off-
1
white solid (75.1 g, 75%): Mp 228-229 °C; Rf = 0.4 (EtOAc-hexanes, 1:1 and 1% of TEA); H
NMR (500 MHz, DMSO-d ) δ 10.64 (s, 1H), 8.57 (dd, J = 4.7, 0.7 Hz, 1H), 8.05 (d, J = 7.9 Hz,
6
1
H), 7.95 (td, J = 7.7, 1.7 Hz, 1H), 7.71 (dd, J = 8.7, 2.3 Hz, 1H), 7.51 (ddd, J = 7.5, 4.8, 1.1 Hz,
1
3
1
H), 7.43 (d, J = 2.3 Hz, 1H), 7.18 (d, J = 8.7 Hz, 1H), 4.23 (s, 2H); C NMR (126 MHz, DMSO-
d6) δ 170.36, 168.13, 156.32, 148.88, 139.31, 137.60, 134.43, 134.17, 127.93, 125.41, 123.93,
1
23.57, 114.50, 57.58; HRMS (ESI/IT-TOF): m/z [M + H]+ calcd for C H BrN O: 316.0080;
14 11 3
found: 316.0076.
Ethyl 8-bromo-6-(pyridin-2-yl)-4H-benzo[f]imidazo[1,5-a][1,4]diazepine-3-carboxylate (4).
A mixture of 7-bromo-5-(pyridin-2-yl)-1,3-dihydro-2H-benzo[e][1,4]diazepin-2-one (3, 90.5 g,
o
2
86.2 mmol) and tetrahydrofuran (1200 mL) was cooled to -20 C using a dry ice / IPA bath. A
solution of potassium t-butoxide (41.8 g, 372.1 mmol) and tetrahydrofuran (300 mL) was added
dropwise to the reaction mixture over a 30 min period, while maintaining the temperature at -20
o
to -15 C. Upon completion of the addition, the reaction mixture was allowed to stir for an
o
additional 60 min at -20 C. Diethyl chlorophosphate (57.9 mL, 400.7 mmol) was then added
27
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 28 of 55
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
o
dropwise to the reaction mixture over 15 min while maintaining the temperature at -20 to -15 C.
Upon completion of the addition, the reaction mixture was allowed to stir for an additional 2 h at
o
-20 C at which point the reaction was deemed complete by analysis by TLC (silica gel and 1%
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
triethylamine / ethyl acetate). Ethyl isocyanoacetate (40.7 mL, 372.1 mmol) was then added
dropwise to the reaction mixture over a 15 min period, while maintaining the temperature at -20
o
to -15 C. Immediately, a solution of potassium t-butoxide (41.8 g, 372.1 mmol) and
tetrahydrofuran (300 mL) was added dropwise to the reaction mixture over 30 min while
o
maintaining the temperature at -20 to -15 C. Upon completion of the addition, the reaction mixture
was allowed to warm to rt and stirred for an additional 12 h at which point the reaction was deemed
complete by analysis by TLC (silica gel and 1% triethylamine / ethyl acetate). The reaction
mixture was then diluted with 5% aq sodium bicarbonate (1000 mL). The biphasic mixture, which
resulted, was allowed to stand for 15 min and the layers were separated. The aq layer was then
extracted twice with dichloromethane (1000 mL x 2) and the combined organic layers were washed
with 5% aq sodium bicarbonate solution (1000 mL) and then 10% aq sodium chloride solution
(
1000 mL). The organic layer was dried (Na SO ). The solvents were removed under reduced
2 4
o
pressure and the residue was slurried with t-butyl methyl ether (1000 mL) at 50 – 55 C for 30 min.
Upon cooling to rt and after holding for 12 h, the solid was filtered and washed with t-butyl methyl
ether (100 mL x 3). The solid was then purified by stirring with ethanol (300 mL) at reflux for 1h.
o
o
Upon cooling to -20 C and after holding at -20 C for 12 h, the solid was filtered off and washed
o
with cold ethanol (50 mL x 2). The solid was dried under vacuum at 40 C to afford the product 4
1
as a light brown powder (61.0 g, 51.8%): MP 212-213 °C; Rf = 0.4 (EtOAc with 1% TEA); H
NMR (500 MHz, CDCl ) δ 8.57 (dd, J = 4.9, 0.9 Hz, 1H), 8.09 (d, J = 8.0 Hz, 1H), 7.89 (s, 1H),
3
7.81 (td, J = 7.8, 1.8 Hz, 1H), 7.78 (dd, J = 8.6, 2.2 Hz, 1H), 7.59 (d, J = 2.2 Hz, 1H), 7.47 (d, J =
28
ACS Paragon Plus Environment

Page 29 of 55
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
8
.6 Hz, 1H), 7.37 (ddd, J = 7.6, 4.8, 1.2 Hz, 1H), 6.12 (d, J = 12.6 Hz, 1H), 4.51 – 4.36 (m, 2H),
1
3
4
.15 (d, J = 12.6 Hz, 1H), 1.43 (t, J = 7.1 Hz, 3H); C NMR (126 MHz, CDCl ) δ 167.08, 162.93,
3
156.20, 148.73, 138.43, 136.94, 135.32, 134.99, 134.57, 134.44, 129.37, 128.54, 124.90, 124.29,
23.99, 120.57, 60.81, 45.02, 14.45; HRMS (ESI/IT-TOF): m/z [M + H]+ calcd for
C H BrN O : 411.0451; found: 411.0454.
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
1
9
16
4
2
Ethyl-6-(pyridin-2-yl)-8-((triisopropylsilyl)ethynyl)-4H-benzo[f]imidazo[1,5-a]
1,4]diazepine-3-carboxylate (5b). A mixture of palladium acetate (1.67 g, 7.4 mmol), tri-o-
[
tolylphosphine (4.51g, 14.8 mmol) and acetonitrile (400 mL) was stirred for 30 min at rt. To the
reaction mixture was then added in sequence ethyl 8-bromo-6-(pyridin-2-yl)-4H-
benzo[f]imidazo[1,5-a][1,4]diazepine-3-carboxylate (4, 61.0 g, 148 mmol), triethylamine (41.3
mL, 297 mmol), (triisopropylsilyl)acetylene (39.9 mL, 178 mmol) and additional acetonitrile (500
o
mL). The reaction mixture was then heated to reflux (75 C) and held for 4 h at which point the
reaction was deemed complete by analysis by TLC (silica gel and 1% triethylamine / ethyl acetate).
Upon completion of the reaction, the mixture was cooled to rt and silica gel (25 g) was added.
After stirring for 30 min, the spent silica gel was removed by filtration and washed with acetonitrile
(100 mL x 2). The solvents were removed under reduced pressure and the residue was dissolved
in dichloromethane (900 mL) and 5% aq sodium bicarbonate (900 mL). The biphasic mixture,
which resulted, was allowed to stand for 15 min and the layers were separated. The aq layer was
then extracted with dichloromethane (900 mL) and the combined organic layers were washed with
5
% aq sodium bicarbonate solution (900 mL) and then 10% aq sodium chloride solution (900 mL).
The organic layer was dried (Na SO ). The solvents were removed under reduced pressure and
2
4
the residue was purified by flash chromatography using silica gel (750 g) and 50% ethyl acetate /
29
ACS Paragon Plus Environment