Structure-Activity Relationship, Drug Metabolism and Pharmacokinetics Properties Optimization, and in Vivo Studies of New Brain Penetrant Triple T-Type Calcium Channel Blockers

Article abstract of DOI:10.1021/acs.jmedchem.6b01356

Despite the availability of numerous antiepileptic drugs, 20-30% of epileptic patients are pharmacoresistant with seizures not appropriately controlled. Consequently, new strategies to address this unmet medical need are required. T-type calcium channels play a key role in neuronal excitability and burst firing, and selective triple T-type calcium channel blockers could offer a new way to treat various CNS disorders, in particular epilepsy. Herein we describe the identification of new 1,4-benzodiazepines as brain penetrant and selective triple T-type calcium channel blockers. From racemic hit 4, optimization work led to the preparation of pyridodiazepine 31c with improved physicochemical properties, solubility, and metabolic stability. The racemic mixture was separated by chiral preparative HPLC, and the resulting lead compound (3R,5S)-31c showed promising efficacy in the WAG/Rij-rat model of generalized nonconvulsive absence-like epilepsy.

Full text of DOI:10.1021/acs.jmedchem.6b01356

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Structure Activity Relationship, Drug Metabolism and
Pharmacokinetics Properties Optimization and In Vivo Studies
of New Brain Penetrant Triple T-Type Calcium Channel Blockers
Romain Siegrist, Davide Pozzi, Gaël Jacob, Caterina Torrisi, Kilian Colas,
Bertrand Braibant, Jacques Mawet, Thomas Pfeifer, Ruben de Kanter,
Catherine Roch, Melanie Kessler, Olivier Corminboeuf, and Olivier Bezençon
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01356 • Publication Date (Web): 04 Nov 2016
Downloaded from http://pubs.acs.org on November 5, 2016
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure Activity Relationship, Drug Metabolism and
Pharmacokinetics Properties Optimization and In Vivo
Studies of New Brain Penetrant Triple TꢀType Calcium
Channel Blockers
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Romain Siegrist‡, Davide Pozzi‡, Gaël Jacob, Caterina Torrisi, Kilian Colas, Bertrand Braibant,
†
Jacques Mawet, Thomas Pfeifer , Ruben de Kanter, Catherine Roch, Melanie Kessler, Olivier
Corminboeuf and Olivier Bezençon*
Drug Discovery Chemistry, Biology and Pharmacology, Actelion Pharmaceuticals Ltd., Gewerbestrasse
1
6, CHꢀ4123 Allschwil, Switzerland.
Tꢀtype calcium channel, Ca 3.1, Ca 3.2, Ca 3.3, 1,4ꢀbenzodiazepines, epilepsy.
v
v
v
ABSTRACT
Despite the availability of numerous antiepileptic drugs, 20ꢀ30% of epileptic patients are
pharmacoresistant with seizures not appropriately controlled. Consequently, new strategies to
address this unmet medical need are required. Tꢀtype calcium channels play a key role in
neuronal excitability and burst firing and selective triple Tꢀtype calcium channel blockers could
offer a new way to treat various CNS disorders, in particular epilepsy. Herein we describe the
identification of new 1,4ꢀbenzodiazepines as brain penetrant and selective triple Tꢀtype calcium
channel blockers. From racemic hit 4, optimization work led to the preparation of
pyridodiazepine 31c with improved physicochemical properties, solubility and metabolic
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 53
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
stability. The racemic mixture was separated by chiral preparative HPLC and the resulting lead
compound (3R,5S)-31c showed promising efficacy in the WAG/Rijꢀrat model of generalized
nonꢀconvulsive absenceꢀlike epilepsy.
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
INTRODUCTION
Epilepsy affects more than fifty million people worldwide. This chronic neurological disorder of
the brain is characterized by an enduring predisposition to generate seizures. More than forty
different epilepsy syndromes, defined by seizure types and other clinical features, are reported.
Around twentyꢀeight antiepileptic drugs (AEDs) are available to patients, many of which
modulate excitatory or inhibitory synaptic activity and have multiple targets. Despite the plethora
of AEDs available and their use in combination therapy, 20ꢀ30% of the epileptic patients are
unable to achieve seizure freedom. Clinical and experimental data indicate that drugꢀresistance
1
might arise from alterations of drug targets or augmentation of multidrug transporter function .
Furthermore, most AEDs present significant side effects such as sedation, cognitive
dysfunctions, weight disorder, or teratogenic potential. Therefore, newer, safer, and more
2
ꢀ3
efficient medications are necessary to treat this unmet medical need .
Calcium channels control many important physiological processes. They allow the influx of
calcium into the cytosol, which lead to the modulation of several cellular processes, like neuronal
excitability, neurotransmitters and hormones release, pain sensation, muscle contractions, or cell
development and proliferation. Both ligandꢀbound or voltageꢀgated calcium channels are known.
Voltageꢀgated channels can be further classified in highꢀ (HVA) or lowꢀvoltageꢀactivated (LVA)
calcium channels. HVA calcium channels are subdivided in Lꢀ, Nꢀ, Pꢀ/Qꢀ, or Rꢀtypes based on
their αꢀsubunits. LVA calcium channels, also referred to as Tꢀtype (T for transient) calcium
ACS Paragon Plus Environment

Page 3 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
channels, are subcategorized in Ca 3.1 (α1G), Ca 3.2 (α1H), and Ca 3.3(α1I). Physiologically,
V
V
V
calcium Tꢀchannels are involved in the regulation of membrane potential and in the control of
4
intracellular calcium concentration . In the brain, the three Tꢀtype calcium channel subtypes are
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
widely expressed and play a key role in the regulation of neuronal excitability and network
6
, 7
oscillation activities .
8
Inhibition of the Tꢀtype calcium channels could be useful for the treatment of several central or
9
10
11
12
13ꢀ
peripheral diseases like Parkinson’s disease , pain , oncology , sleep disorders , or epilepsy
1
5
16
and could also play a role in neuroprotection . Strong evidences link Tꢀtype calcium channels
with epilepsy. Pharmacological studies using genetic models of generalized nonꢀconvulsive
absenceꢀlike epilepsy (the generalized absence epilepsy rat of Strasbourg,GAERS or the
WAG/Rij (Wistar Albino Glaxo from Rijswijkꢀrat) showed that Tꢀtype calcium channel blockers
1
7, 18
strongly decreased the epileptic activity in those models
. In addition, mice overexpressing
1
9
the Ca 3.1 channels present frequent spike wave discharges (the hallmark of absence seizures)
V
20
whereas, knockout mice lacking the Ca 3.1 gene are protected from absence seizures . Finally,
V
mutations in CACNA1H, the gene encoding the Ca 3.2 channel, have been found in patients
V
2
1
suffering from idiopathic generalized epilepsy .
Despite the attractive therapeutic potential, more than twenty years were necessary to overcome
the poor specificity of the antihypertensive drug mibefradil, a Tꢀ and Lꢀtype calcium channel
2
2
blocker, and to discover the first selective Tꢀtype calcium channel blocker . Since then, several
chemically diverse selective triple Tꢀtype calcium channel blockers (blocking the 3 subtypes)
2
3, 24
have been reported
and at least three of them have reached clinical trials (Figure 1). Efforts
at Merck on different chemical classes led to the discovery of the brainꢀpenetrant and selective
2
5a
triple blocker MKꢀ8998 (1) , which displayed single digit nanomolar activities on the three Tꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 53
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
25
type calcium channels . This compound was advanced up to a phase II clinical trial for the
treatment of acute psychosis in patients with schizophrenia. In this study, 1 was ineffective.
However, olanzapine, the active comparator used in this study, displayed only a limited efficacy
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
26
and therefore the utility of Tꢀchannel antagonists to treat schizophrenia cannot be ruled out .
1
8
Zalicus (now Epirus) discovered Zꢀ944 (2) , which successfully completed a phase Ib clinical
trial. Interestingly, (R)ꢀ2ꢀ(4ꢀcyclopropylphenyl)ꢀNꢀ(1ꢀ(5ꢀ(2,2,2ꢀtrifluoroethoxy)pyridinꢀ2ꢀ
yl)ethyl)acetamide (TTAꢀA2) , a close analog of 1, and 2 were both assessed and efficacious in
2
5
1
8, 27
rodent models of seizure and epilepsy
. Recently, AbbVie reported the discontinuation of
28
their program with the peripherally acting selective triple blocker ABTꢀ639 (3) for the
treatment of diabetic neuropathies. Additionally, several known antiepileptic drugs, like
ethosuximide, phenytoin, valproic acid or zonisamide are claimed to exert at least part of their
13, 29
efficacy through Tꢀtype channel inhibition
.
Figure 1. Triple T-type Calcium Channel Blockers that Have Been Studied in the Clinic
Our effort started with an HTS campaign, using a highꢀthroughput calcium flux assay (FLIPR)
3
0
on cells expressing the recombinant human channels as previously reported . The screening led
ACS Paragon Plus Environment

Page 5 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
to the identification of hit 4 and its close analog 5 (Figure 2). Whereas benzodiazepine 4
exhibited a promising biological activity towards the Ca 3.1, Ca 3.2 and Ca 3.3 channels, a
V
V
V
good selectivity over the Ca 1.2 channel, and an acceptable intellectual property position, the
V
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
compound suffered from a poor solubility, a high lipophilicity, and a low metabolic stability. A
metabolite identification was performed to locate soft spots on the molecule. A 10 ꢀM solution
of 5 in buffer was incubated for 30 min in the presence of human liver microsomes (HLM) and
the formation of two major metabolites was observed (Figure 3). Thus, 5 was debenzylated or
oxidized on the benzodiazepine scaffold, leading to the formation of metabolites 6 or 7,
respectively.
Figure 2. Structures of Hits 4 and 5
Figure 3. Structure of Major Metabolites 6 and 7 Observed upon Incubation with HLM
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 53
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 this work, we present the optimization of racemic hit 4 by reducing its lipophilicity and
improving its metabolic stability as well as its solubility resulting in lead compound (3R,5S)-31c
which showed promising efficacy in the WAG/Rijꢀrat model of generalized nonꢀconvulsive
absenceꢀlike epilepsy.
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
CHEMISTRY
In order to quickly study the structure activity relationship (SAR) around the urea moiety, the
replacement of the benzyl group and the substitution of the benzodiazepine scaffold, we
developed an efficient synthesis toward the two key building blocks 12 and 13. As depicted in
Scheme 1, our synthesis started with an amide coupling between aminoacetophenone (8) and
racemic Bocꢀalanine. A brief survey of activating reagent (EDC•HCl, BTFFH, HATU,…)
revealed that the best yields to prepare amide 9 were obtained using POCl in pyridine.
3
Substituted aminoacetophenone were also tolerated, allowing the introduction of substituents
onto the phenyl of the benzodiazepine core (vide infra). The Boc protecting group was cleaved
under acidic conditions and the resulting primary amine underwent an intramolecular
condensation with the ketone moiety affording imine 10. The diastereoselectivity of the imine
reduction was easily tuned by the choice of the right reaction conditions. Thus, whereas
reduction with sodium borohydride led to a 1:1 mixture of the cisꢀ and transꢀisomers, an
hydrogenation afforded almost exclusively the cisꢀisomer (cis-11). Notably, the first column
chromatography of this synthetic sequence was performed after the imine reduction. Lactam cis-
11 was then reduced with borane, and the secondary, aliphatic amine was selectively Bocꢀ
protected, leading to the preparation of our first key intermediate 12. Alkylation with benzyl
bromide and subsequent Bocꢀdeprotection yielded the targeted secondary amine 13.
ACS Paragon Plus Environment

Page 7 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 1. Synthesis of the Benzodiazepine Core (Amide Strategy)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) (±)ꢀBocAlaOH, POCl , pyridine, 0 °C to rt, 2 h, 96%; (b) TFA,
3
DCM, 0 °C to rt, 18 h, 62%; (c) Pd/C, H , MeOH, rt, 48 h, 87%; (d) BH3
•DMS, THF, reflux, 18
2
h, 56%; (e) Boc O, Cs CO , DCM, rt, 3 d, 84%; (f) BnBr, Cs CO , NaI, DMF, 110 °C, 2 h, 95%;
2
2
3
2
3
(
g) 4M HCl in dioxane, DCM, rt, 6 h, quantitative.
Starting from secondary amine 13, various urea 14a-d were synthesized either through reaction
with commercially available isocyanates, or from primary amines which were activated with
3
1
bistrifluoroethoxy carbonate . Amides 15a-d were conveniently synthesized from the
corresponding carboxylic acids using HATU, EDC HCl, or POCl as activating agents (See
•
3
supporting information for details). Best results were however obtained using HATU as coupling
reagent, as exemplified for the synthesis of 15d. Finally, sulfonamide 16 was obtained through
reaction with the corresponding sulfonyl chloride (Scheme 2).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 53
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
a
Scheme 2. Urea Variation or Replacement
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
Reagents and conditions: (a) RꢀNCO, NEt , MeCN, rt, 18 h, 24ꢀ81%; (b) RꢀNH ,
3
2
(
F CCH O) CO, DIPEA, MeCN, 75 °C, 2 h, 23%; (c) morpholinꢀ4ꢀyl acetic acid, HATU, NEt ,
3 2 2 3
DCM, rt, 18 h, 63%; (d) 1ꢀmethylꢀ1Hꢀimidazoleꢀ4ꢀsulfonyl chloride, NEt , DCM, rt, 18 h, 53%.
3
In order to replace the benzyl moiety, we first alkylated intermediate 12 to obtain compound 17
(
Scheme 3). Boc deprotection and subsequent urea formation yielded final compounds 19.
Whereas the alkylation worked nicely under harsh conditions with substituted benzylbromides,
erratic yields were obtained with aliphaticꢀ or etherꢀhalides. We eventually discovered that
variation could easily by introduced through reductive amination between intermediates 12 or
3
2
18 and an aldehyde. As all desired aldehydes could not be synthesized, we turned our attention
3
3
to the methodology reported by Sorribes et al. . Thus, in the presence of the Karstedt’s catalyst,
carboxylic acids were suitable substrates to alkylate intermediates 12. Alternatively, intermediate
12 was subjected to BuchwaldꢀHartwig cross coupling, acylation, or treated with a sulfonyl
ACS Paragon Plus Environment

Page 9 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
chloride to give, after Boc deprotection and urea formation, final products 19v, 19w, or 19x
respectively (see Table 3).
a
Scheme 3. Benzyl Variation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
1
1
Reagents and conditions: (a) R ꢀBr, DIPEA, DMF, 180 °C (ꢀW), 15 min, 26ꢀ90%; (b) R ꢀ
1
CHO, NaBH(OAc) , DCE, 50 °C, 2 h, 6ꢀ95%; (c) R ꢀCO H, PhSiH , dppe, [Pt] Karstedt’s
3
2
3
catalyst, nBu O, 60 °C, 18 h, 59%; (d) TFA, DCM, 0 °C to rt, 18 h, 97%ꢀquantitative; (e)
2
BnNCO or pyridineꢀ3ꢀisocyanate, NEt , DCM or MeCN, 0 °C, 1 h, 13ꢀ95%.
3
Synthesis of phenyl substituted benzodiazepines 20-25 (Table 4) were performed either from the
corresponding substituted aminoacetophenone as previously mentioned and described in Scheme
3
2
1
and 2, or by late state functionalization of bromide 21 . Thus, cyanation of the aryl halide
34ꢀ38
following Rosenmundꢀ von Braun
protocol led to nitrile derivative 22, while treatment with
3
9
trifluoroethanol and NaH in presence of CuBr2 gave the corresponding ether 23 (Scheme 4).
Alternatively, a Negishi coupling allowed for introduction of a methyl group (24) whereas a
Lꢀ
4
0, 41
prolineꢀpromoted CuIꢀcatalyzed coupling between 21 and sodium methanesulfinate
aryl sulfone 25.
gave
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 53
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
a
Scheme 4. Phenyl Ring Substitution
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
Reagents and conditions: (a) CuBr , NaH, CF CH OH, DMF, 110 °C, 18 h, 58%; (b) ZnCN,
2
3
2
Pd dba , dppf, PMHS, DMAC, 150 °C (ꢀW), 40 min, 36%; (c) NaSO Me,
L
ꢀProline, CuI,
2
3
2
DMSO, 110 °C, 48 h, 47%; (d) Me Zn, Pd(PPh ) , THF, 50 °C, overnight, 68%; (e) 4M HCl
2
3 4
dioxane, rt, overnight, 22 (quantitative), 23 (quantitative), 24 (quantitative), 25 (quantitative); (f)
pyridineꢀ3ꢀisocyanate, NEt , MeCN, overnight, 22 (71%), 23 (64%), 24 (84%), 25 (85%).
3
Apart from the classical amide couplingꢀcyclizationꢀreduction strategy, a second approach for
the construction of the pyridodiazepinic core structure is represented in Scheme 5a and involves
an aromatic nuclophilic substitution (S Ar) as a key step. The synthesis of derivative 31c is
N
4
2
reported as example. Reaction between 3ꢀfluoroisonicotinonitrile (26) and primary amine 27
led to pyridine 28, which was in turn converted into the corresponding methyl ketone upon
treatment with MeLi in THF. Boc deprotection led to imine 29, which was susbequently
benzylated on the anilic nitrogen using NaH as a base and benzyl bromide in DMF at 70 °C.
ACS Paragon Plus Environment

Page 11 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hydrogenation of the imine with Pd/C yielded 30 with a cis/trans diastereoselectivity of 9:1. The
two diastereoisomers could be easily separated by column chromatography and cis-30 was
submitted to urea formation under standard conditions to afford 31c. Alternatively, the alkyl side
chain could be introduced from the beginning of the sequence. For example secondary amine
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
3
33 was reacted with 3,5ꢀdifluoroisonicotinonitrile (32) to give 34 (Scheme 5b). The synthesis
continued as described for 31c to afford 31l in a 5 steps sequence. Interestingly, hydrogenation
of the imine with Pd/C was found to be ineffective when F was present, while changing the
catalyst to PtO led to the formation of the antiꢀstereoisomer as the major product (cis/trans 4:6).
2
Finally, reduction with sodium triacetoxyborohydride resulted in a 1:1 cis/trans ratio. These two
approaches were not limited to the synthesis of pyridyldiazepines only but could also be applied
for the synthesis of some phenyl substituted benzodiazepines.
a
Scheme 5. Synthesis of the Benzodiazepine Pyridine Core (S Ar Strategy)
N
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 53
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
a
Reagents and conditions: (a) NEt , NMP, 90 °C, overnight, 85%; (b) MeLi, THF, 0 °C to rt,
3
overnight, 52%; (c) 4M HCl in dioxane, rt, overnight, 100%; (d) NaH, BnBr, DMF, 70 °C, 2 h,
0%; (e) Pd/C, H , MeOH, rt, 90 min, 77%; (f) BnNCO, DIPEA DCM, rt, overnight, 80%; (g)
6
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
NMP, DIPEA 120 °C, overnight, 69%; (h) MeLi, THF, ꢀ78 °C to 0 °, overnight, 25%; (i) 4M
HCl in dioxane, rt, 1 h, 98%; (k) NaBH(OAc) , MeOH, rt, 1 h, 27% cisꢀisomer, 27% transꢀ
3
isomer; (l) BnNCO, DIPEA DCM, rt, overnight, 34%.
RESULTS AND DISCUSSION
The optimization work started by assessing the importance of the two cisꢀmethyl groups on the
benzodiazepine core. To this end, a methyl scan was performed (Table 1). Removal of one or of
both methyl groups led to a significant loss in potency, as could be observed for compounds 35a,
3
5b, or 35c. Displacing a methyl from the 3ꢀ to the 2ꢀposition of the benzodiazepine scaffold was
also detrimental and both cisꢀ or transꢀanalogs 35d or 35e were not further pursued. The
corresponding transꢀisomer of initial hit (4), 35f, was also clearly less active. Finally, switching
to a benzodiazepinone scaffold, as exemplified with 35g or 36, was not well tolerated either. In
conclusion, the two methyl groups in a cisꢀconformation were required to achieve a suitable
biological potency. Importantly, derivatives 35a-g and 36 were all found to be metabolically
unstable.
Next, optically pure enantiomers (3S,5R)-4 and (3R,5S)-4 were synthesized. A chiral resolution
of racemic intermediate cisꢀ17 using a column charged with a chiral stationary phase afforded
enantiomers (3S,5R)-17 and (3R,5S)-17 (Scheme 6). The Boc protecting group was cleaved
under acidic conditions to give optically pure secondary amines (3S,5R)-13 and (3R,5S)-13
ACS Paragon Plus Environment

Page 13 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
which were converted into final compounds (3S,5R)-4 or (3R,5S)-4 upon treatment with benzyl
isocyanate. Enantiomer (3R,5S)-4 was found more potent than the other one. To determine the
absolute configuration of the most potent enantiomer, amine (3S,5R)-13 (precursor of the less
active enantiomer) was converted into biologically inactive Mosher’s amide (3S,5R)-37 upon
treatment with (S)ꢀ(+)ꢀαꢀmethoxyꢀαꢀtrifuloromethylphenylacetyl chloride. By slow diffusion of
pentane into a concentrated solution of (3S,5R)-37 in diethyl ether, suitable single crystals were
obtained to carry out an Xꢀray crystal structure analysis, leading to the conclusion that the most
active enantiomer presented the (3R,5S) dimethyl conformation. For the sake of convenience,
most of the optimization work was however performed using racemic material (see below).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Core Modification: Importance of the two Methyl Groups
Ca 3.1
Clint
Ca 3.1
Clint
v
v
Ca 3.2
HLM
Ca 3.2
HLM
v
v
Cmpd.
R
Cmpd.
R
Ca 3.3
(µL/min/
Ca 3.3 IC
(µL/min/
v
v
50
a
b
a
b
IC (nM)
mg)
(nM)
mg)
5
0
17 ± 7
35 ± 25
8 ± 4
279 ± 55
4 (Ref.)
ꢀꢀꢀ
>1250
35f
340 ± 340
95 ± 34
>1250
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 53
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
93 ± 26
304 ± 109
682 ± 266
259 ± 132
3
5a
5b
241 ± 98
>1250
>1250
35g
>1250
>1250
1
1
64 ± 51
02 ± 19
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
676 ± 292
618 ± 334
1589 ± 696
3
141 ± 25
6 ± 9
36
ꢀꢀꢀ
2
3
05 ± 58
262 ± 74
358 ± 138
641 ± 101
3
5c
(3S,5R)-
3
92 ± 158
>1250
n.d.
n.d.
n.d.
4
2
14 ± 77
93 ± 27
2
5 ± 1
17 ± 10
2 ± 0.1
(
3R,5S)-
3
5d
464 ± 248
003 ± 144
4
1
2
11± 31
198 ± 88
26 ± 279
3
5e
n.d.
3
a
b
Values represent the average of at least three independent runs. n= 1.
a
Scheme 6. Synthesis of Mosher’s Amide (3S,5R)-37
ACS Paragon Plus Environment

Page 15 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
Boc
Boc
Boc
N
H
N
N
N
N
a
b,c
N
N
N
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
±)
cis-17
(3S,5R)-17
(3R,5S)-17
(3R,5S)-4
b
O
N
H
NH
N
d
c
N
N
(3S,5R)-13
(3S,5R)-4
(3S,5R)-37
a
Reagents and conditions: (a) Prep. chiral HPLC (column (R,R) WhelkꢀO1,
heptane:EtOH/0.1%DEA 1:9), (3S,5R)-17 (46%), (3R,5S)-17 (48%); (b) 4M HCl in dioxane,
EtOH, rt, 2 d, quantitative; (c) BnNCO, NEt , MeCN, rt, 18 h, 65ꢀ76%; (d) (S)ꢀ(+)ꢀαꢀmethoxyꢀ
3
αꢀtrifuloromethylphenylacetyl chloride, DIPEA, DCM, rt, 2 d, 45%.
Next, modification of the benzyl urea moiety (Table 2) demonstrated that this region of the
molecule is fairly tolerant. Indeed, several ureas 14a-d or amides 15a-d led to potent
compounds. Notably, polar groups like an oxetane or pyridine were suitable substrates and
helped to tune the physicochemical properties of the final compounds (vide infra). Alkylation of
the urea moiety resulted in loss of activity, as can be seen for compound 38. Sulfonamides or
carbamates were clearly less potent, as exemplified with 16 or 39. Finally, a carbonyl function
was absolutely mandatory to retain a good biological activity, as tertiary basic amine like
derivative 40 were inactive.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 53
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
Whereas several amides or ureas proved to be well tolerated and afforded potent triple blockage,
these replacements did not lead to significant improvement in the intrinsic clearance. However,
solubility of the compounds was markedly enhanced by the introduction of an heterocycle or an
ether moiety and compounds 14b, 14c, or 15c exhibited a solubility ranging from 21 to 39
ꢀg/mL at pH 7. The most interesting compound identified during the replacement of the benzyl
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
urea was oxetane analog 14d, for which both solubility (148 ꢀg/mL at pH 7) and human
metabolic stability (539 ꢀL/min/mg) were significantly improved.
Table 2: Variation of the Urea Moiety
Ca 3.1
Clint
HLM
Ca 3.1
Clint
HLM
v
v
Ca 3.2
Ca 3.2
v
v
1
1
Cmpd.
R
Cmpd.
R
Ca 3.3
(µL/min/
Ca 3.3 IC
(µL/min/
v
v
50
a
b
a
b
IC (nM)
mg)
(nM)
mg)
5
0
1
7 ± 7
34 ± 4
55 ± 25
6 ± 1.6
36 ± 18
51 ± 15
7 ± 5
4
(Ref.)
-
35 ± 25
>1250
>1250
15c
15d
>1250
>1250
8
8
± 4
± 2
1
4a
14 ± 1.5
± 2
6
ACS Paragon Plus Environment

Page 17 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
2 ± 24
576 ± 6
1490 ± 11
210 ± 9
1
4b
61 ± 24
>1250
>1250
16
38
n.d.
n.d.
1
2
4 ± 3
6 ± 4
457 ± 212
835 ± 403
892 ± 255
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
4c
5
8 ± 13
± 3
6 ± 9
61 ± 20
2 ± 2
9
2
420 ± 114
841 ± 200
1
4d
539
>1250
1160
39
40
n.d.
n.d
1
325 473± 27
31 ± 4
38 ± 18
10 ± 3
6110 ± 13.4
7410 ± 47
15a
>10000 ± 9
2
4 ± 9
47 ± 31
± 4
1
5b
8
a
b
Values represent the average of at least three independent runs. n=1.
As mentioned earlier, debenzylation was identified as the main mechanism of metabolism.
Several replacements were therefore introduced to block this soft spot. Substitution of the benzyl
moiety with one (compounds 19a, 19b, or 19c) or two fluorine atoms (compounds 19d or 19e)
led to very potent triple Tꢀtype calcium channel blockers. Replacement of the phenyl moiety
with unsubstituted pyridines resulted in less active analogs (e.g. 19f, 19g, or 19h). In the case of
2
ꢀ or 3ꢀpyridines, the activity was recovered by the introduction of a suitable substituent (see 19i,
1
9j, or 19k). The most potent triple Tꢀtype calcium channel blockers of this work were identified
by the introduction of substituted 5ꢀmembered heterocycles, like furane derivatives 19l or 19m,
or thiazoles 19n or to a lesser extent 19o. Bulky alkyl chains or even cycloalkyl groups could
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 53
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
also be tolerated whereas less hindered alkyl chains showed lower potencies. Thus, whereas
ethyl derivative 19p was not very potent, isopentyl 19q or cyclopentyl 19r displayed an
interesting biological activity. Finally, we introduced ether moieties, as exemplified by
derivatives 19s, 19t, or 19u. Notably, the introduction of this more polar moieties resulted in a
remarkable improvement in solubility (248 ꢀg/mL at pH 7 for 19s or 81 ꢀg/mL at pH 7 for 19t).
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
Whereas several suitable replacements for the benzyl groups led to potent compounds and
enhanced in some cases the solubility, none of these were able to prevent metabolic degradation.
We therefore removed the ꢀCH ꢀ linker, but the resulting phenyl analog 19v was metabolically
2
unstable. Introduction of an amide (19w) or a sulfonamide (19x) resulted in a massive loss in
4
4
potency. Attempts to hinder the metabolism by deuterium introduction (19y) were also
unsuccessful. Finally, only the inactive debenzylated analog 18 proved to be metabolically stable
when incubated with HLM.
Table 3. Replacement of the Benzyl Moiety
Ca 3.1
Clint
Ca 3.1
v
v
Clint
Ca 3.2
HLM
Ca 3.2
v
v
HLM
1
2
1
2
Cmpd.
R
R
Ca 3.3
(µL/mi Cmpd.
n
R
R
Ca 3.3
v
v
(
µL/min
IC50
IC50
c
/
mg)
a
c
a
(
nM)
/mg)
(nM)
ACS Paragon Plus Environment

Page 19 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
7 ± 7
13 ± 6
15 ± 6
6 ± 2
4
ꢀꢀꢀ
ꢀꢀꢀ
35 ± 25
>1250
1200
19n
19o
Bn
>1250
>1250
(
Ref.)
8
4
± 4
± 1
33 ± 7
42 ± 9
33 ± 18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
9a
3ꢀPy
11 ± 2.5
3ꢀPy
5
± 3
7
54 ±
25
22 ±
2
8
± 2
9
1
9b
3ꢀPy
14 ± 3
5 ± 2
>1250
19p
Bn
>1250
346
139 ±
56
2
5 ± 6
23 ± 14
91 ± 53
12 ± 7
1
9c
3ꢀPy
3ꢀPy
31 ± 7
7 ± 2
>1250
>1250
19q
19r
3ꢀPy
3ꢀPy
>1250
1150
1
3 ± 2
23 ± 5
62 ± 13
14 ± 4
1
9d
18 ± 14
6
± 2
± 1
8
75 ±
b
3
25
5
2
670 ±
1
9e
Bn
9 ± 1
n.d.
19s
3ꢀPy
720
b
980
4 ± 0.3
243 ±
b
77
605 ±
216
108 ± 12
91 ± 20
23 ± 8
19f
Bn
n.d.
19t
3ꢀPy
1230
479 ±
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 53
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
2
74
7 ± 26
76 ±
46
500 ±
06
9
4
2
35 ± 17
44 ± 12
12 ± 3
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
9g
Bn
Bn
Bn
n.d.
19u
19v
19w
Bn
Bn
Bn
n.d.
>1250
n.d.
2
9
5 ± 6
b
>
7260 ±
3
874
2
56 ± 20
209 ± 55
225 ± 73
b
>8580 ±
3705
1
9h
n.d.
b
1710 ±
650
6070 ±
b
25 ± 5
42 ± 8
14 ± 6
400
19i
>1250
>11550 ±
b
1250
b
>10000
b
>10000
b
53 ± 20
49 ± 1
4695 ±
19j
Bn
Bn
>1250
>1250
19x
19y
3ꢀPy
3ꢀPy
1985
n.d.
b
1
0 ± 2
1810 ±
3
50
4
5 ± 5
27
1
9k
33 ± 5
0 ± 3
25
4
1110
1
ACS Paragon Plus Environment

Page 21 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
b
b
7
± 1.4
>10000
>10700
1
9l
Bn
13 ± 2
>1250
>1250
18
Bn
59
b
6
± 1
± 1
>8605
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
9m
3ꢀPy
14 ± 5
± 6
9
a
b
c
Values represent the average of at least three independent runs unless otherwise stated. n=2.
n=1.
After modification of the urea and replacement of the benzyl moiety we concentrated our efforts
on the benzodiazepinic scaffold with the following goals: keep a high potency on the Ca 3.1,
V
Ca 3.2, and Ca 3.3 Tꢀtype channels, decrease the lipophilicity of the molecule and block the
V
V
metabolically weak spots. First, we studied the effect of the substitution of the phenyl ring on
activity and on human liver microsomal stability (Table 4). In this regard fluorine resulted to be
1
the best substituent. In particular, incorporation of a fluorine atom at position R (20a) allowed to
keep comparable activity to reference compound 4 while moderately improving HLM intrinsic
2
3
clearance. On the other hand, fluorine substitution at R (20b), or R (20c) gave potent
4
compounds that remained unstable in HLM. Finally, introduction of a fluorine at position R
(20d) allowed to improve the HLM intrinsic clearance but resulted in a considerable decrease in
the IC values. The substitution with more polar or bulkier groups such as ꢀOMe, ꢀCN, ꢀSO Me,
5
0
2
ꢀ
OCH CF led to an important loss in potency even though in some cases improved HLM values
2 3
were observed (20g, 20h and 22). In general, the substitution of the benzodiazepinic phenyl ring
showed to be sensitive and provided no substantial improvement of metabolic stability in human
liver microsomes without substantially reducing activity. Furthermore, in none of the examples
the rat liver microsomal intrinsic clearance (RLM) could be improved (RLM >1250 µL/min/mg)
.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 53
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 4: Activities and Microsomal Stability of Phenyl Substituted Benzodiazepines
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
Ca 3.1
Clint
HLM
Ca 3.1
Clint
HLM
v
v
Ca 3.2
Ca 3.2
v
v
Cmpd.
Scaffold
Cmpd.
Scaffold
Ca 3.3 IC
(µL/min
Ca 3.3 IC
(µL/min/
v
50
v
50
a
c
a
c
(
nM)
/mg)
(nM)
mg)
1
7 ± 7
35 ± 25
± 4
627 ± 460
1600 ± 961
610 ± 201
4
ꢀꢀꢀ
>1250
20h
649
n.d.
(
Ref.)
8
2
0 ± 9
52 ± 18
0 ± 3
>5300
>6500
>9000
2
0a
628
20i
1
68 ± 29
112 ± 35
8 ± 3
370 ± 46
414 ± 111
73 ± 27
20b
20c
>1250
>1250
20j
n.d.
n.d.
6 ± 1
286 ± 21
428 ± 43
20k
12 ± 4
ACS Paragon Plus Environment

Page 23 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
± 1
137 ± 81
b
1
78 ± 37
582 ± 4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b
2
0d
205 ± 137
720
22
860 ± 391
418
>1250
n.d.
b
1
1
7 ± 5
2 ± 2
308 ± 84
1
87 ± 104
2
0e
873
23
790 ± 289
22 ± 14
8 ± 5
3557 ± 1690
188 ± 94
305 ± 93
130 ± 69
227 ± 21
231 ± 64
74 ± 54
20f
n.d.
521
24
25
390 ± 80
530 ± 125
100 ± 29
>10000
>8000
20g
n.d
>10000
a
b
c
Values represent the average of at least three independent runs unless otherwise stated. n= 2.
n= 1.
Driven by the observation that a fluorine atom on the phenyl ring could moderately stabilize the
molecule, we further investigated this effect on the so far best combinations (Table 5). If on one
hand the substitution with a fluorine led to an important lowering of the HLM values for 41, on
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 53
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
the other hand it was able to only moderately stabilize compounds showing already interesting in
vitro clearance (43 vs. 42, 45 vs. 44 and 46 vs. 14d).
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
Table 5: Effect of Fluorine Substitution on HLM
Cmpd.
R1
R2
R3
Ca 3.1
Clint
v
Ca 3.2
HLM
v
Ca 3.3
(µL/min/
v
a
b
IC (nM)
mg)
5
0
1
7 ± 7
35 ± 25
± 4
1 ± 4
8 ± 18
4
(Ref.)
ꢀꢀꢀ
ꢀꢀꢀ
ꢀꢀꢀ
>1250
>1250
8
3
1
5a
H
3
1
0 ± 3
7 ± 15
2 ± 11
4
5
4
1
2
F
562
412
17 ± 5
2
9 ± 14
62 ± 22
13 ± 3
4
H
ACS Paragon Plus Environment

Page 25 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
5 ± 21
66 ± 18
2 ± 13
0 ± 12
30 ± 10
3 ± 5
15 ± 53
93 ± 40
4
4
4
3
4
5
F
H
F
357
332
193
539
375
5
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
1
2
2
0 ± 8
6 ± 9
1
4d
H
F
61 ± 20
12 ± 2
40 ± 15
59 ± 14
13 ± 6
46
a
b
Values represent the average of at least three independent runs. n= 1.
With the goal to further stabilize our derivatives against phase I metabolism, we modified the
benzodiazepinic scaffold replacing the phenyl ring portion by sixꢀ and five membered ring
heterocycles (Table 6). Pyridodiazepine derivatives 31a and 31b proved to be only moderately
active or inactive compared to 4 respectively, while 31c and 31d showed interesting activity on
the three channels. Importantly, 31c presented an encouraging in vitro clearance of 276
µL/min/mg. We therefore studied the effect of the substitution of these two pyridodiazepine
derivatives on HLM stability and on activity on the Tꢀtype calcium channels. For the
pyridodiazepines with N at 9ꢀposition (31d), the substitution in ortho with –Cl (31e), –F (31g),
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 53
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
and –Me (31i) did improve neither the already excellent activity nor the metabolic stability,
whereas the substitution with a –CF led to a moderate improvement in HLM stability but also to
3
a loss in potency (31j). The introduction of substituents at 7ꢀposition (31f and 31h) could
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
moderately stabilize the molecule while keeping acceptable IC values. Nevertheless, the most
5
0
interesting results were obtained with pyridodiazepines with N at 8ꢀposition. In particular 31l
where a fluorine atom was introduced in meta position displayed the lowest HLM stability while
keeping an acceptable potency on the Ca 3.1, ꢀ3.2 and ꢀ3.3 Tꢀtype channels. Lastly,
v
pyrimidodiazepine 31n and pyrazolodiazepine 31o were inactive on the target. However, in none
of the examples the RLM stability could be improved by introducing an heterocycle (RLM
>1250 µL/min/mg).
With 31c and 31l in our hands, we investigated the replacement of the benzyl moiety and the
modification of the urea part (Table 7). In contrast to the 1,4ꢀbenzodiazepines with the naked
scaffold, 1,4ꢀpyridodiazepines were less active upon ureas replacement with amides and at the
same time the HLM stability dropped considerably (e.g. 47). Compared to parent benzodiazepine
31c, bis deuterated benzodiazepine 48 showed a slightly improved metabolic stability in
presence of HLM. Introduction of a trifluoromethylfurane moiety instead of a benzyl group
allowed to improve the potency (52) while keeping the HLM stability constant, whereas the
introduction of o,mꢀdifluorobenzylꢀ or oꢀfluorobenzylꢀsubstituents allowed at the same time to
boost the activity and to improve further the HLM stability (49 and 50).
ACS Paragon Plus Environment

Page 27 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 6: Benzodiazepinic Core Modification by Incorporation of Heterocycles
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ca 3.1
Ca 3.1
Clint
v
v
Clint HLM
Ca 3.2
Ca 3.2
HLM
v
v
Cmpd.
Scaffold
(µL/min/
Cmpd.
Scaffold
Ca 3.3 IC
Ca 3.3
(µL/min/
v
50
v
c
mg)
a
a
c
(
nM)
IC (nM)
mg)
5
0
1
7 ± 7
35 ± 25
± 4
65 ± 12
95 ± 9
15 ± 6
4
(Ref.)
ꢀꢀꢀ
>1250
31h
837
8
5
97 ± 29
12 ± 23
1
8 ± 5
28 ± 4
3 ± 9
2
3
1a
n.d
31i
1240
9
1 ± 24
3
6
280 ±
b
1
670
57 ± 15
104 ± 11
281 ± 78
b
b
3
1b
>11000
8740
n.d.
31j
531
>
1
2
42 ± 52
36 ± 89
2
20 ± 20
224 ± 52
5 ± 19
3
1c
276
31k
>1250
2
5 ± 10
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 53
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
5 ± 3
1 ± 16
± 3
1
40 ± 9
72 ± 11
± 7
3
3
1d
1060
1005
409
31l
31m
31n
31o
185
n.d.
n.d
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
2
3
4
5
6
7
8
9
0
1
139 ± 333
95 ± 346
06 ± 29
1
6 ± 5
6
3
1e
36 ± 7
2
2
2 ± 12
>
8500
7
8 ± 10
>6400
3
1f
68 ± 25
46 ± 18
1000 ± 316
b
>9400
7 ± 1
6 ± 1
b
>7300
31g
>1250
n.d.
b
>10000
4 ± 0.2
a
b
c
Values represent the average of at least three independent runs unless otherwise stated. n=2 .
n= 1.
Interestingly, the pꢀmethoxybenzyl urea improved both the activity and the intrinsic clearance in
human liver microsomes at once as shown in 51. Keeping the urea part fixed and replacing the
aromatic benzyl group by an isopentyl allowed to significantly improve the metabolic stability
in HLM without observing a drop in the activity (54). The combination of the pꢀmethoxybenzyl
urea with the trifluoromethylfurane led to benzodiazepine 53 which presented an excellent
activity on all three Tꢀtype calcium channels, similar stability in presence of HLM but also
improved stability in RLM compared to 51. Similarly, in derivative 55 the introduction of the 3ꢀ
ACS Paragon Plus Environment

Page 29 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pyridyl urea allowed to ameliorate the HLM stability by a factor 3 and also to improve
considerably the RLM stability. Hindering the metabolism by the introduction of deuterium at
the benzylic position, allowed to improve slightly the metabolic stability in HLM (56). Once
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
again the replacement of the benzyl substituent by an isopentyl group (58) allowed to stabilize
the molecule in human liver microsomes by a factor 2 and the introduction of a fluorine
substituent meta to the N atom allowed to improve the IC , especially of the Ca 3.2 Tꢀchannel
5
0
v
type (57). Nonetheless, beside the clear drop in activity, an important drawback of 3ꢀpyridyl
ureas 55, 57 and 58 was the low brain penetration potential across the blood brain barrier (BBB).
Indeed, we measured a set of pyridodiazepines in the MDR1/MDCK assay which allows
evaluation of passive permeability and potential interaction with human drug efflux Pꢀ
glycoprotein transporter (Figure 4). Derivative 31c showed an acceptable value for the efflux
ratio (Papp basolateral to apical (BA)/ P app apical to basolateral (AB)) of 3.2 and a good passive
ꢀ6
permeability (Papp (AB) = 16.6 (10 cm/s)) indicating that the compound is a weak substrate for
human Pꢀgp and has the potential to cross the BBB. In 31l, the introduction of a fluorine atom
allowed to lower the pyridine’s basic pKa by two logunits (from 6.2 to 4.2) resulting in an
ꢀ6
improved efflux ratio of 2.3 with good passive permeability Papp (AB) = 26.3 (10 cm/s).
Nonetheless, replacement of the benzyl urea by the 3ꢀpyridyl urea resulted in a benzodiazepine
(57) showing a high efflux ratio of 60 suggesting that this compound is a strong substrate of
human Pꢀgp. The similarity in the logD values between 31c and 57 (3.1 and 3.3 respectively)
suggests that the introduction of a second basic center with a pKa≅ 4.4 is detrimental for the
permeation.
ACS Paragon Plus Environment