Discovery of epigenetic regulator i-bet762: Lead optimization to afford a clinical candidate inhibitor of the bet bromodomains

Article abstract of DOI:10.1021/jm401088k

The bromo and extra C-terminal domain (BET) family of bromodomains are involved in binding epigenetic marks on histone proteins, more specifically acetylated lysine residues. This paper describes the discovery and structure-activity relationships (SAR) of potent benzodiazepine inhibitors that disrupt the function of the BET family of bromodomains (BRD2, BRD3, and BRD4). This work has yielded a potent, selective compound I-BET762 that is now under evaluation in a phase I/II clinical trial for nuclear protein in testis (NUT) midline carcinoma and other cancers.

Full text of DOI:10.1021/jm401088k

Subscriber access provided by Temple University Libraries
Article
Discovery of epigenetic regulator I-BET762: lead optimization to
afford a clinical candidate inhibitor of the BET bromodomains
Olivier Mirguet, Romain Gosmini, Jérôme Toum, Catherine-Anne Clement, Mélanie Barnathan, Jean-
Marie Brusq, Jacqueline E Mordaunt, Richard Grimes, Miriam Crowe, Olivier Pineau, Myriam Ajakane,
Alain Daugan, Phillip Jeffrey, Leanne Cutler, Andrea Haynes, Nicholas Smithers, Chun-Wa Chung, Paul
Bamborough, Iain J Uings, Tony Lewis, Jason Witherington, Nigel Parr, Rab Prinjha, and Edwige Nicodeme
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm401088k • Publication Date (Web): 09 Sep 2013
Downloaded from http://pubs.acs.org on September 15, 2013
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 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Witherington, Jason; Epinova DPU, Immuno-Inflammation Therapeutic
Area, GlaxoSmithKline,
Parr, Nigel; Epinova DPU, Immuno-Inflammation Therapeutic Area,
GlaxoSmithKline,
Prinjha, Rab; Epinova DPU, Immuno-Inflammation Therapeutic Area,
GlaxoSmithKline,
Nicodeme, Edwige; Centre de Recherches François Hyafil, GlaxoSmithKline
R&D,
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Discovery of epigenetic regulator I-BET762: lead
optimization to afford a clinical candidate inhibitor
of the BET bromodomains
1
Olivier Mirguet, † Romain Gosmini, Jérôme Toum, Catherine A. Clément, Mélanie
,
,*
1, ¥
1
1
1
1
3
3
Barnathan, Jean-marie Brusq, Jacqueline E. Mordaunt, Richard M. Grimes, Miriam Crowe,
3
1
Olivier Pineau, Myriam Ajakane, Alain Daugan, Phillip Jeffrey, Leanne Cutler, Andrea C.
1
1
4
5
5
4
2
2
4
Haynes, Nicholas N. Smithers, Chun-wa Chung, Paul Bamborough, Iain J. Uings, Antonia
3
Lewis, Jason Witherington, Nigel Parr, Rab K. Prinjha, Edwige Nicodème
4
4
4
1
1
2
Centre de Recherches François Hyafil, GlaxoSmithKline R&D, 25 avenue du Québec. 91140 Villebon-sur-Yvette, France.
Computational & Structural Chemistry, Molecular Discovery Research, GlaxoSmithKline R&D, Medicines Research Centre, Gunnels Wood
Road, Stevenage, Hertfordshire SG1 2NY, U.K.
Screening & Compound Profiling, Molecular Discovery Research, GlaxoSmithKline R&D, Medicines Research Centre, Gunnels Wood Road,
3
Stevenage, Hertfordshire SG1 2NY, U.K.
4 Epinova, Discovery Performance Unit, GlaxoSmithKline R&D, Stevenage, Hertfordshire SG1 2NY, U.K.
5
Quantitative Pharmacology, EMU Discovery Performance Unit,GlaxoSmithKline R&D, Stevenage, Hertfordshire SG1 2NY, U.K.
*Corresponding Author: E-mail address: oliviermirguet@yahoo.fr
KEYWORDS. BET inhibitors, benzodiazepine, I-BET, I-BET762, GSK525762, bromodomains,
epigenetics.
ACS Paragon Plus Environment
1

Page 3 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT. The bromo and extra C-terminal domain (BET) family of bromodomains are
involved in binding epigenetic marks on histone proteins, more specifically acetylated lysine
residues. This paper describes the discovery and structure-activity relationship (SAR) of potent
benzodiazepine inhibitors that disrupt the function of the BET family of bromodomains (BRD2,
BRD3 and BRD4). This work has yielded a potent, selective compound I-BET762 that is now
under evaluation in a phase I/II clinical trial for nuclear protein in testis (NUT) midline
carcinoma and other cancers.
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
Introduction
The bromodomain and extra-terminal domain (BET) family of proteins (BRD2, BRD3, BRD4
and BRDT) activate transcription through their ability to recognize specific ε-N-acetyl-modified
1
lysine residues found within histone tails and other proteins. While three members of this family
(BRD2, BRD3 and BRD4) are ubiquitously expressed, the fourth member, BRDT, is a tissue-
restricted, chromatin-associated protein expressed in pachytene spermatocytes, diplotene
2
spermatocytes, and round spermatids. BRD2-4 function as chromatin scaffolds that recruit
3
Polymerase II (RNA Pol II) to the PAFc complexes, and in the case of BRD4 P-TEFb, thus
ensuring RNA Pol II-dependent transcriptional initiation and elongation respectively.
The recent disclosure of potent, selective small molecule inhibitors of the BET family of
bromodomains demonstrate that epigenetic reader proteins may be as tractable to small molecule
4
drug discovery as their epigenetic enzyme counterparts. To date, a small number of structurally
diverse chemotypes have been reported to target the acetyl-binding pocket of the BRD domains
of the BET proteins and demonstrate a broad spectrum of desirable biological effects (Chart 1).
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 60
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
5
Of these, I-BET762 1 has recently entered clinical trials for NUT midline carcinoma (NMC).
6
7
NMC is a rare, lethal and aggressive tumor with a median overall survival of 6.7 months, in
which a NUT gene translocation generates a fusion protein with BRD4 or BRD3 proteins that is
retained strictly in the cell nucleus via interactions with chromatin. Functional studies in patient-
derived NMC cell lines have validated the essential role of the BRD4–NUT fusion oncoprotein
in maintaining the characteristic proliferation advantage and differentiation block of this
malignancy which is arrested following treatment with the thienodiazepine BET inhibitor JQ1
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
(compound 2). JQ1 is also active against myeloma, lymphoma, acute lymphoblastic
9
10
1
leukemia and neuroblastoma in vitro and in vivo, while a second class of quinoline based
1
12
BET protein inhibitors, exemplified by I-BET151 (3), was shown to have considerable pre-
1
clinical activity against acute leukemia including MLL-related acute myeloid leukemia. More
3
14
15
recently, publications from Pfizer/SGC and the Chinese Academy of Sciences described 3-
methyl-1,2,3,4-tetrahydroquinazolin-2-ones 4 and 2-thiazolidinones 5 respectively, as new
chemical probes for BET protein inhibition.
ACS Paragon Plus Environment
3

Page 5 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
Chart 1. Reported structurally diverse BET inhibitors.
We recently reported the discovery of compounds using a reporter assay for ApoA1
upregulation, whose molecular targets were subsequently elucidated to be the BET family of
16
bromodomain-containing proteins. A recent disclosure from the Canadian-based biotechnology
company Resverlogix has reported the mechanism of action for its Phase IIb asset RVX-208 6
4c
Chart 1), which was developed as a regulator of ApoA1 gene transcription, to be also through
(
inhibition of the BET bromodomain modules. This provides increased confidence in the clinical
tolerability of these novel agents and highlights the value of understanding the SAR around
inhibitors of this protein family to fully exploit and test the therapeutic potential of this target
class.
In this manuscript we describe the medicinal chemistry program that led to the discovery of the
potent and selective small molecule I-BET762 or GSK525762 which is currently under clinical
development.
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 60
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
RESULTS AND DISCUSSIONS
Discovery of small molecule upregulators of ApoA1
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
Efforts to target endogenous upregulation of ApoA1 expression are hampered by the absence
of an obvious molecular mechanism around which to build a drug discovery effort. In 2001, we
therefore generated a stable human HepG2 hepatocyte cell-line containing an ApoA1 luciferase
reporter and used this to screen compounds in order to identify molecules with the ability to
upregulate reporter gene activity. Diversity and targeted screening approaches led to the
identification of the benzodiazepine (BZD) 7 (Figure 1) which showed potent induction of the
1
7
ApoA1 reporter gene with an EC170 of 0.22 µM.
Figure 1. Benzodiazepine hit 7.
Compound 7 was considered a suitable starting point for lead optimization. In the absence of any
knowledge of the molecular target at this time, a program of medicinal chemistry was carried out
16
to optimize the potency against the ApoA1 upregulation assay. Having subsequently identified
the targets of our compounds as the bromo and extra-terminal (BET) family of bromodomains,
ACS Paragon Plus Environment
5

Page 7 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
we will describe their affinities towards all BRD subtypes and the lead optimization program that
led to the discovery of the clinical compound I-BET762 (1).
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
Chemistry
The 1,4-benzodiazepine skeleton is one of medicinal chemistry’s most widely used scaffolds.
1
8
First discovered in the 1960s, there are now many marketed pharmaceutical agents (e.g.
Bromozepam) that contain a 1,4-benzodiazepine motif principally as GABA receptor positive
allosteric modulators for CNS clinical indications, and a number of efficient synthetic routes to
3-amino-1,4-benzodiazepines have been described in the literature. The design of novel ApoA1
upregulators was performed to explore all different chemical modifications around the BZD
scaffold as depicted in Schemes 1 to 7. 3-Amino-1,4-benzodiazepine derivatives 7, 8 and 9a-c
19, 20
described in this manuscript were prepared using the synthetic route described in Scheme 1.
21
Commercially available benzodiazepinone 7 was first submitted to catalytic hydrogenation and
the Cbz group was removed leading to amine 8 as a key intermediate to explore diversity at
3
position R . Since compound 7 tolerates a polar carbamate linker, we explored other spacers such
as amide 9a, urea 9b, sulfonamide 9c using standard chemistries.
a
Scheme 1. Synthesis of substituted 3-amino-1,4-benzodiazepine derivatives 7, 8 and 9a-c
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
3
Reagents and Conditions: (a) 1,4-cyclohexadiene, cat. Pd/C, MeOH, RT, 98%; (b) R
introduction, see Experimental Section, 2 – 70%.
As depicted in Scheme 2, BZD 13 bearing an acetamide linker was obtained starting from
2
2
known alprazolam analogue 10. Deprotonation of the BZD methylene with LiHMDS in THF at
low temperature followed by addition of ethyl bromoacetate led to alkylated product 11 in
modest yield. The latter was saponified using sodium hydroxide giving acid 12 in good yield.
Finally, standard peptide coupling conditions with HATU as coupling reagent afforded BZD
acetamide 13 in 71% yield.
a
Scheme 2. Synthesis of benzodiazepine 13
a
Reagents and Conditions: (a) Ethyl bromoacetate, LiHMDS, THF, -78°C to RT, 48%; (b) 1N
NaOH, EtOH/THF, RT, 78%; (c) Benzylamine, HATU, Et N, DMF, RT, 71%.
3
As depicted in Scheme 3, 3-amino-1,4-benzodiazepines 16a-h were synthesized with good
31
overall yields using a reported sequence starting from aminobenzophenones 14a-h and
ACS Paragon Plus Environment
7

Page 9 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
23
Katritzky reagent 15. Thionation with Lawesson’s reagent afforded thioamides 17a-h and
2
triazole ring construction using conditions previously described led to test compounds 18a-h.
1
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
Scheme 3. Representative synthesis of substituted 3-amino-1,4-benzodiazepine derivatives
a
Reagents and Conditions: (a) 15, EDCI, cat. DMAP, DCM, 0°C; (b) NH , MeOH, RT; (c)
3
NH OAc, AcOH, RT; (d) Lawesson’s reagent, toluene, reflux; (e) H N-NH •H O, MeOH, RT;
4
2
2
2
(f) MeC(OMe) , PPTS, MeOH, reflux; Bt = benzotriazole.
3
Exploration of the carbamate substitution was performed using the chemistry shown in Scheme
4. Intermediate 8 was reacted with a variety of formate derivatives to afford the desired
carbamates 19a-c.
a
Scheme 4. Synthesis of substituted 3-amino-1,4-benzodiazepine derivatives 19a-c
a
3
Reagents and Conditions: (a) R introduction, see Experimental Section, 10 – 62%.
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 60
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
Enantiomers of compound 19c were separated by chiral preparative HPLC using a Chiralpak
AD column and an isocratic mixture of EtOH/heptane: 80/20 (Scheme 5). The absolute
stereochemistries of (+)-19c and (-)-19c were assigned using Vibrational Circular Dichroism (see
Supplementary Information Section 2.2 for details) and found to be R and S respectively.
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
Scheme 5. Chiral separation of compound 19c
a
19c
(+)-19c
(-)-19c
a
Reagents and Conditions: (a) chiral preparative HPLC (Chiralpak AD column, EtOH/heptane
80/20).
:
Alternatively, compound 19d was prepared using chemistry depicted in Scheme 6.
24
Benzodiazepinone 20 was subjected to acid hydrolysis to give amine 21 in good yield. Upon
reaction with ethyl chloroformate in the presence of triethylamine, compound 22 was obtained
from 21. Thioamide 23 was prepared using the same chemistry described earlier and, finally,
formation of the triazole ring was performed in three steps involving addition of hydrazine
followed by a quench with acetylchloride. Cyclization occurred under refluxing acid conditions
giving test compound 19d in good yield over the three steps.
a
Scheme 6. Synthesis of substituted 3-amino-1,4-benzodiazepine derivatives 19d
ACS Paragon Plus Environment
9

Page 11 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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) 37% HBr, AcOH, 80°C, 92%; (b) ethyl chloroformate, Et N,
3
THF, 0°C to RT, 83%; (c) Lawesson’s reagent, toluene, reflux; (d) H N-NH •H O, MeOH, RT;
2
2
2
(
e) AcCl, DIPEA, THF, 0°C; (f) AcOH, reflux, 55% over 3 steps.
Acetamide BZD derivatives were prepared according to Scheme 7. As described in our
1
6
previous report, the thionation step of the synthetic route to racemic final compounds involved
the use of Lawesson’s reagent. Under those conditions, thioamide 25 was obtained as a racemic
mixture. In order to obtain the final compounds as active pure (S)-enantiomers, we separated
enantiomers by preparative chiral HPLC and tested them to identify the active samples, which
proved to be arduous. We thus investigated the thionation step and found that using phosphorus
pentasulfide in the presence of sodium carbonate led to the formation of the desired product
without epimerization. The completion of the synthesis was consistent with that described earlier
2
and afforded compounds 26, 27 and 28a-g. Compounds 29 and 30 were prepared using the
5
same synthetic pathway.
a
Scheme 7. Synthesis of BZD acetamide derivatives 26, 27 and 28a-g
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and Conditions: (a) P S , Na CO , 1,2-DCE, 65°C; (b) H N-NH , THF, 5°C to
2
4
10
2
3
2
3
5°C; (c) AcCl, THF, 0°C; (d) AcOH, THF, RT; (e) NaOH, THF, RT; (f) R XH, coupling
1
reagent, base, solvent (X= NH or O).
Biological activities
1
6
We earlier reported that ApoA1 upregulation in HepG2 cells tracks in vitro binding to the three
subtypes of BET proteins extremely well. Although lead optimization was carried out against
ApoA1 upregulation as the primary assay, in this manuscript for clarity we will discuss only the
BET affinities of the benzodiazepine compounds. ApoA1 upregulation data for all compounds is
fully detailed in the Supporting Information section (Tables S1 to S4).
Compound 7 was found to be around micro molar IC in our fluorescence polarization assay
5
0
(Figure 1) against each of the three BET isoforms. This is more potent than simpler
2
benzodiazepine analogues lacking the benzyl carbamate group, such as Alprazolam or the
6
36
analogous thienotriazolodiazepine GABA receptor positive allosteric modulator Etizolam
(Figure 2A). The crystal structures of these simple unsubstituted thienotriazolodiazepines in
complex with the N-terminal bromodomain of BRD4 were solved and found to adopt a binding
1
mode perfectly superimposable with that of 7 previously described. In these structures, the
6
ACS Paragon Plus Environment
11

Page 13 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
triazolyl ring forms close hydrogen-bonding interactions with the side chain NH of Asn140 and a
network of water molecules (Figure 2). The BZD fused phenyl ring fills a narrow lipophilic
channel between Leu92 and Pro82, termed the “ZA channel”, and the pendant phenyl ring packs
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
16
against Ile146, on the “WPF shelf”.
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2: A) GABA receptor positive allosteric modulator Etizolam and its analogue 31.
Crystal structures of BRD4 in complex with B) Etizolam, C) 31, D) 7 (the benzyl carbamate
substituent depicted has good density within this structure but places the benzyl group in an
ACS Paragon Plus Environment
13

Page 15 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
orientation distinct from that seen in BRD2 suggesting this group may be mobile), E) 7 showing
the electrostatic potential surface of BRD4, F) Comparative binding to BRD4 of compound 1
(yellow carbons) and the acetyl-lysine sidechain of a histone H4 peptide (white carbons, PDB
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
4c
entry 2yel).
In its complex with BRD4, the additional carbamate group of 7 extends towards the volume
that would be occupied by the N-terminal side of the acetyl-lysine peptide chain. The carbamate
NH makes hydrogen-bonding interactions to the side-chain of Asn140, however no further
specific interactions are observed for the rest of the carbamate. The carbamate benzyl group is
found to adopt distinct conformations in BRD2 and BRD4, again suggestive that this group does
not make strong interactions with the protein and are mobile. A similar conclusion was reached
26
by Filippakopoulos et al. in their studies comparing Alprazolam and Midazolam to (+)-JQ1.
We decided to explore alternative functionalities to the carbamate (Table 1) capable of
maintaining the interaction of the NH with Asn140. We removed it completely (compound 8,
Table 1) or replaced it with an amide (compound 9a), an urea (compound 9b) and a sulfonamide
(
compound 9c). We also investigated linkers where the NH group is moved further from the
benzodiazepine core, for example the acetamide compound 13. Urea 9b displays similar binding
affinities to 7, but amide 9a and sulfonamide 9c were less active. We attribute this to steric non-
complementarity between the bromodomain site and benzyl substituents when attached via
linkers whose preferred geometry is different to the linear carbamate and urea. This hypothesis
has been confirmed by X-ray crystallography (data not shown).
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 60
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
Compounds were also tested for their anti-inflammatory properties assessing their ability to
inhibit IL-6 production after a LPS-challenge in peripheral blood mononuclear cells (PBMC). In
this cellular assay, compound 9b was found to be much weaker than 7, probably due to lower
solubility of the urea compared to carbamate. Overall, the carbamate substituent looked the most
attractive on the benzodiazepine ring 3-position.
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
Table 1. SAR around the BZD ring RHS substituent.
BRD2 (FP)
pIC50
BRD3 (FP)
pIC50
BRD4 (FP)
pIC50
PBMC (IL-6)
pIC50
Cmpd
R
7
NHCO
2
Bn
5.9
6.2
6.3
6.8
c
c
c
b
8
NH
2
4.6
4.9
4.7
NT
b
9a
NHCOCH
2
CH
2
Ph
Ph
Ph
CONHCH Ph
4.8
5.7
4.5
5.5
6.1
5.2
5.7
5.4
5.9
5.2
5.2
NT
9
b
NHCONHCH
NHSO CH
CH
2
6.4
4.8
5.3
9
c
2
2
c, d
1
3
2
2
5.6
a
b
c
d
All results are means of n ≥ 2 unless stated. NT = Not Tested. n = 1. Following repeat
testing, activity was observed but no pIC was defined on six separate test occasions due to low
max response observed (<50% inhibition) at the highest concentration tested.
5
0
ACS Paragon Plus Environment
15

Page 17 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
We next investigated the substituents on the fused phenyl ring of the benzodiazepine (Table 2).
This aromatic ring binds into the lipophilic ZA channel of the bromodomain (Figure 2).
Introduction of electron-withdrawing (compounds 18a and 18b) or donating groups (compounds
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
8c and 18d) at either position 8 or 9 did not have much effect on affinities toward the 3 subunits
except for the 8-nitro derivative 18a which lost quite some potency on the bromodomain and in
the cell assay. The other substituents were well tolerated and the 8-methoxy compound 18d was
found to be the best, especially taking into account its potency within the cell assay.
We then turned our attention toward the pendant phenyl ring that interacts with the lipophilic
WPF shelf (Figure 2) and first tested all positions of the ring. Introduction of a methoxy at the
ortho position of the ring abolished BRD affinity, as exemplified by compound 18e. This can be
rationalised with reference to the X-ray structure of Etizolam (Figure 2B). The ortho-chloro
substituent induces a rotation in the inter-ring torsion angle, leading to a decrease in the surface
area of the pendant phenyl ring in contact with the hydrophobic surface of Ile146 on the WPF
shelf. When the methoxy substituent was introduced at either meta or para position, compounds
1
8f and 18g respectively were found equally potent against the 3 BRD subunits and in the cell
assay, probably due to the fact that those two positions point toward the solvent and would not
make direct interactions with the bromodomain. The para position of the pendant phenyl ring
tolerates most substitutions. For example, a methoxy group (compound 18g) could be replaced
by a methyl (compound 18h) without affecting potency.
a
Table 2. SAR around both phenyl rings of the BZD.
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Central
BZDR
pIC50
1
R , R
2
BRD2 (FP)
pIC50
BRD3 (FP)
pIC50
BRD4 (FP) PBMC (IL-6)
pIC50
Cmpd
pIC50
7
H, H
5.9
6.2
5.5
5.8
6.3
6.8
6.2
b
1
8a
8b
8-NO
2
, H
4.9
5.5
5.5
5.9
5.7
6.2
6.2
6.9
5.0
6.3
6.5
NT
c
1
8-Cl, H
7.3
c
c
c
c
18c
8d
18e
8f
9-Me, H
8-OMe, H
H, 2’-OMe
H, 3’-OMe
H, 4’-OMe
H, 4’-Me
5.6
5.9
5.8
6.1
b
1
6.1
5.0
5.5
5.7
5.8
6.1
5.3
6.0
6.0
6.3
5.3
6.0
5.9
5.9
NT
b
NT
c
1
<5.0
c
1
8g
8h
<5.0
b
c
1
6.0
NT
<5.0
a
b
c
All results are means of n ≥ 2 unless stated. NT = Not Tested. n = 1.
As discussed earlier, the 1,4-benzodiazepine motif has been used principally as GABA
receptor positive allosteric modulators for CNS clinical indications. From the start of this work,
we recognized the structural similarity between our benzodiazepine ApoA1 upregulators and
2
alprazolam , and thus tested the ability of our compounds to inhibit diazepam binding to the
6
central GABA receptor. Compounds 7, 18b and 18c were found active in this assay with a pIC50
ACS Paragon Plus Environment
17

Page 19 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 7.3 for 18b. The hypothesis developed by Filippakopoulos et al. that JQ1 has no significant
GABA receptor activity due to its bulky substitution at the 3 position of the benzodiazepine ring
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
2
6
system is inconsistent with our observations, since all our compounds share a similar bulky
group but are still active on the central GABA receptor. Nonetheless, when a substituent was
introduced at either meta or para position of the pendant phenyl ring, compounds 18f, 18g and
1
8h were found to be inactive on the GABA receptor, thereby giving us an opportunity to solve
this selectivity issue.
Having defined that the carbamate function was the best linker at position 3 of the
benzodiazepine ring, we turned our attention to optimizing this substituent. From the crystal
structure of 7 in BRD4, this group points toward the bulk solvent, and provides limited potential
to improve bromodomain affinity. However, optimization of this part of the molecule allowed us
to manipulate the physicochemical properties of the compounds (Table 3). Introduction of a
fluorine atom at the para position of the benzyl group (compound 19a) maintained BRD
affinities and cellular activity. Our attempt to truncate the benzyl group by a 3-fluorophenyl
(
compound 19b) resulted in reduced affinities toward all three bromodomains. We then tried to
reduce both molecular weight (MW) and lipophilicity (clogP) of the molecule to a range more
desirable for oral drugs (MW <400, clogP <3) by replacing the benzyl group by an alkyl such as
in compound 19c. This had a similar affinity to 7 against both BRD3 and BRD4 although there
was a slight drop in cellular potency. We then investigated the contribution of each enantiomer to
the BRD affinity of compound 19c and showed that only enantiomer (+)-19c was found active.
Ultimately this proved to be of absolute (R)-configuration, whereas its enantiomer (-)-19c had no
significant activity.
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 60
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
Table 3. SAR around the BZD carbamate substituent.
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
d
c
BRD2 (FP) BRD3 (FP) BRD4 (FP) PBMC (IL-6) MW
t1/2
2
R , R
3
clogP
2.4
Cmpd
pIC50
pIC50
pIC50
pIC50
h
7
H, Bn
H, 4F-Bn
H, 3F-Ph
H, Et
5.9
6.2
6.3
6.8
423
441
427
361
361
361
0.58
1
9a
19b
9c
-)-19c
+)-19c
9d
5.7
6
6.3
6.5
2.5
0.33
e
e
e
b
b
5.5
5.9
5.8
NT
3.1
NT
1
5.6
4
6
4
6.1
4
6.3
2.1
0.56
b
b
(
H, Et
NT
2.1
NT
(
H, Et
5.9
5.6
6.3
5.9
6.4
6.1
6.7
2.1
0.23
0.75
b
1
4’-Cl, Et
NT
396
2.8
a
b
c
All results are means of n ≥ 2 unless stated. NT = Not Tested. calculated logP
d
(Biobyte/Daylight). t determined in pH 2 buffer by HPLC. n = 1.
e
1
/2
As discussed earlier, we speculated that compounds unsubstituted on the pendant phenyl ring
such as 19a-c would be active on the GABA receptor. In order to eliminate this undesired
activity, we introduced a substituent in the para position of the pendant ring. Compound 19d was
equipotent to 19c and 7 on the BRD subtypes but completely inactive when tested in the central
ACS Paragon Plus Environment
19

Page 21 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
GABA receptor binding assay (pIC50 < 5), thus reinforcing our belief that this pendant ring
substitution was key to GABA receptor selectivity of the benzodiazepine compounds.
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
When discussing the progression of such compounds, we speculated that the presence of two
nitrogens on the benzodiazepine 4-carbon could lead to instability of the compounds. It has been
shown that triazolobenzo- and triazolothienodiazepines can undergo a ring-opening reaction in
27
an acidic aqueous solution, leading to a postulated intermediate 32 in the case of compound
(+)-19c as shown in Scheme 8. The latter could easily be hydrolyzed in acidic conditions to form
the benzophenone 33. This instability at low pH could lead to a rapid degradation of (+)-19c
which was incompatible with our target of achieving good oral exposure. We tried to discharge
this concern by setting up an assay where compound half-life could be measured in an acidic
solution that mimicks gastric pH. Compounds were suspended in a pH 2.0 buffer and their half-
life (t ) measured. All compounds tested from this series were found to undergo rapid ring
1/2
opening under these low pH conditions (t1/2 < 1 h, Table 3), with the active enantiomer (+)-19c
being the most rapidly converted. Since this may be a possible route for active metabolite
formation, we turned our attention to structural changes that could mitigate this potential risk.
Scheme 8. Postulated mechanism for a potential low pH degradation pathway
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Additionally, we also wanted to address anticipated issues in achieving an efficient asymmetric
synthesis for compounds of this series. Although there is one report that describes the synthesis
of an enantiomerically pure aminobenzodiazepine using a crystallization-induced asymmetric
28
29
transformation, most reported syntheses involve a chiral resolution step. We tried to solve
those two concerns together by designing a compound that would be acid-stable and amenable to
an enantioselective synthesis. One way of improving acid stability would be to remove one
nitrogen atom at position 3 of the benzodiazepine ring. That would lead to compounds like the
benzodiazepine acetamide 13 discussed earlier, which was found to be acid stable under the
same conditions (t1/2 > 2h). We also recognized that this compound could be synthesized from
natural aspartic acid, and was amenable to an enantioselective synthesis as shown in Scheme 7.
Compound 13 showed only modest affinity toward the BRD subunits (Table 1) but we
speculated that applying SAR learnt during the optimization of the aminobenzodiazepine series
would ultimately lead to improvements. We also knew that compound 13 had no affinity toward
the central BZD receptor (pIC50 < 5.0) which reinforced our interest in this sub-series. The
inactivity on the GABA receptor was true for all compounds tested in this sub-series (see
Supporting Information table S4 for compounds tested).
In order to improve BRD affinity, we investigated the structure-activity relationships on the
amide linker exemplified by 13. The best features from our earlier compounds were
ACS Paragon Plus Environment
21

Page 23 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
incorporated, specifically the methoxy group at the 8 position of the benzodiazepine exemplified
in compound 18d (see Table 2), and a para chlorine atom on the pendant phenyl ring, to yield
compounds shown in Table 4. As seen in the aminobenzodiazepine series, a variety of
substituents were tolerated in this part of the molecule, probably due to the fact that this group is
solvent exposed. Substituted phenyl (compound 28a), heteroaromatic ring (compounds 28b-d)
and alkyl substituted analogues (compounds 28e and 1) were found equipotent toward the three
BRD tested. This position was therefore used to improve the overall physicochemical properties
of the compounds. Aromatic substituents proved to be more lipophilic than desirable (clogP >3)
although this made little impact on the solubility, which was good for all compounds. Alkyl-
substituted analogues (e.g. 28e and 1) offered a better profile regarding lipophilicity, with 1
being one of the most potent acetamide analogues in the cell assay. We then assessed the
importance of the methoxy at position 8 and tested analogue 29 having the methoxy at position 9
of the fused ring of the BZD. This compound is less potent in both BRD binding and cellular
assays, reinforcing the preference for substitution at position 8. When we explored the meta
position of the pendant phenyl ring, compound 30 bearing a fluorine atom was found less active
than 1. Carboxylic acid 27 was also tested and found to be barely active, a result which could be
explained by the likely proximity of the acid to the side-chain carbonyl oxygen of Asn140
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
(Figure 2D). Methyl ester 26 and butyl ester 28g had excellent affinity toward the three BRD
sub-types and a very good potency in the cell assay, probably due to a concordant increase in
their lipophilicity. Ester 28g was the most potent compound we identified in the cell assay, with
a pIC of 7.4. Consistent with other reported BET inhibitors there is no selectivity observed with
5
0
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 60
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
these molecules between BET protein isoforms (Table S4), reflecting the high degree of
1
6a
sequence homology between the acetyl lysine binding sites of these proteins.
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
Table 4. SAR around the BZD acetamide template.
BRD4
(FP)
pIC50
PBMC
(IL-6)
pIC50
Solubility
in FaSSIF
µg/mL
1
2
3
d
Cmpd
R
R
X
R
clogP
2.9
4.2
3.3
3.1
3.1
2.8
2.4
2.4
1.8
2.4
2.8
1
3
H
H
NH
Bn
5.2
6.3
6.4
6.4
6.3
6.3
6.2
5.3
5.7
6.0
6.5
1
2
8a
8b
8c
8d
8e
8-OMe 4’-Cl NH
8-OMe 4’-Cl NH
8-OMe 4’-Cl NH
8-OMe 4’-Cl NH
4F-Ph
105
100
115
95
2
2-thiazole
2-pyridyl
3-pyridyl
2
b
2
NT
b
2
8-OMe 4’-Cl NH c-propylmethyl
NT
111
122
1
8-OMe 4’-Cl NH
9-OMe 4’-Cl NH
ethyl
ethyl
ethyl
H
6.5
6.2
c
b
2
9
0
7
6
5.7
NT
b
3
2
2
8-OMe
3’-F
NH
O
5.9
NT
116
113
115
c
8-OMe 4’-Cl
8-OMe 4’-Cl
5.1
<4
O
methyl
6.6
6.9
ACS Paragon Plus Environment
23

Page 25 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
28g
8-OMe 4’-Cl
O
butyl
6.4
7.4
104
4.4
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
b
c
d
All results are means of n ≥ 2 unless stated. NT = Not Tested. n = 1. calculated logP
(Biobyte/Daylight).
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
From this set of compounds, due to its good overall balance of properties we decided to select
compound 1, also known as I-BET or I-BET762, for further profiling. The X-ray structure of 1
bound to BRD4 (Figure 2F) illustrates an unexpected replacement of the direct hydrogen bond
from the sidechain NH of Asn140 to the carbamate NH of 7 by a water-bridged interaction
through the amide NH of 1. Other binding features of the two complexes are very similar.
I-BET762 Preclinical Developability Profile
I-BET762 has high in vitro passive permeability (167 nm/sec) with similar moderate free
fractions across all the species investigated (f ca. 0.2; Table 5). I-BET762 has high solubility in
ub
physiologically relevant media demonstrating solubility > 3 mg/mL in all vehicles tested
(
Simulated Gastric Fluid (pH 1.6) > 4 mg/mL, fasted state simulated intestinal media (pH 6.5) >
3 mg/mL and fed state simulated intestinal fluid (pH 6.5) > 5 mg/mL). In vitro data suggest that
I-BET762 has a low potential to inhibit the major human CYP isoforms (1A2, 2C9, 2C19, 2D6,
3
A4 IC ’s ≥ 33 ꢀM) with no evidence for time dependent inhibition of CYP2D6 or CYP3A4. In
50
vitro clearances determined with liver microsomes and hepatocytes were low across all species,
including human (CL ≤ 1.7 mL/min/g liver, all systems).
i
Table 5. Summary of in vitro DMPK data for 1
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 60
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
b
Assay
Test system
Blood
Mouse
0.21
Rat
0.18
1.45
<0.86
Dog
0.24
0.69
< 1.7
Primate
Human
0.19
a
c
Free Fraction (fub)
NT
Metabolic stability
Microsomes
Hepatocytes
<0.53
<0.86
0.20
<0.53
<0.86
CL , (mL/min/g liver)
i
<0.085
a
Measured using rapid equilibrium dialysis at a nominal concentration of 1000 ng/mL.
c
Cynomolgus monkey. NT = Not Tested.
b
The pharmacokinetic properties of I-BET762 were then investigated in 4 preclinical species:
mouse, rat, dog and primate (Table 6). I-BET762 has a low blood clearance in dog and primate
(
≤ 30% liver blood flow in both species) and a moderate blood clearance in the mouse and rat (≤
7
0% liver blood flow in both species). The volumes of distribution were moderate (ca. 1.8 L/kg)
in rat, dog and primate, but high in mouse (ca. 6.5 L/kg) indicating distribution into the tissues.
The terminal half-life was short in rat (ca. 0.5 h) and longer in mouse, dog and primate (ca. 1.5 -
5
.9 h, respectively). Oral bioavailability of I-BET762 was moderate (ca. 27%) in the rat and
higher in mouse, dog and primate (44 – 61%).
The solubility and permeability data suggested that I-BET762 should diffuse well across
membranes. However, in vitro metabolic stability data for I-BET762 were found not to predict
the in vivo PK profile across the four pre-clinical species. The favourable pharmacokinetic
ACS Paragon Plus Environment
25

Page 27 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
profile observed in the mouse, dog and primate was not observed in the rat despite broadly
similar in vitro clearance values. The exact reasons for this are not understood, but it is known
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
3
2
that the rat often poorly predicts oral bioavailability in human. Using all the available data,
human pharmacokinetic predictions were performed using several techniques including
allometric scaling and physiologically based pharmacokinetic modelling (PBPK) using
Gastroplus™ (version 5.0; Simulations Plus Inc., Lancaster, CA, USA; data not shown). Despite
the less than optimal pharmacokinetic profile observed in the rat a favourable human
pharmacokinetic profile was predicted thus justifying clinical progression.
a
Table 6. Mouse, rat, dog and primate pharmacokinetic parameters for 1
AUC0-inf
ng.h/mL
468
CLb
mL/min/kg
75
Vd
L/kg
6.5
T , i.v.
1/2
F, p.o.
%
Species
h
Balb/c Mouse
CD Rat
1.5
0.5
5.9
1.5
61
201
4741
3338
63
5
1.8
1.8
1.5
27
44
52
Beagle Dog
b
Primate
14
a
Values are mean, n=3. IV dose (1 mg/kg) was 1 h infusion in DMSO (2%, v/v) Kleptose HPB
10%, w/v) in saline (0.9%, w/v). PO dose (3 mg/kg) suspension in 1% (w/v) methylcellulose
(400 cps) (aq). Cynomolgus monkey. Values are mean, n=4. IV dose (2 mg/kg) was 1 h
infusion in DMSO (4%, v/v) HPB (20%, w/v) in water. PO dose (5 mg/kg) suspension in 0.5%
(
b
(
w/v) HPMC/ TWEEN80 1% in water.
I-BET762 was tested in a non-GLP bacterial mutation screening assay (Ames test) with
Salmonella typhimurium TA1535, TA1537, TA98, TA100 and Escherichia coli
WP2uvrA(pKM101) in the presence and absence of rat S9-mix. I-BET762 is not mutagenic
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 60
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
when tested in either the presence or absence of S9-mix. The compound had no appreciable
activity (pIC /pEC ≤ 4.8) in an in-house developability panel containing 38 assays related to
50
50
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
3
34
safety or an external panel containing 50 selectivity assays (<30% inhibition at 10 ꢀM).
Moreover, representative bromodomain containing proteins BAZ2B, SP140, ATAD2, CREBBP
7
and PCAF were shown not to interact with I-BET.
In vivo studies using I-BET762 have demonstrated efficacy in a range of oncology and
5
, 9, 10
immunoinflammatory models.
Consistent with previously reported effects of I-BET on T-
35
cell function in vitro, I-BET dose-dependently inhibited ear swelling in a rat model of delayed-
type hypersensitivity (DTH) at a comparable level to the positive control rapamycin (Figure 3).
These data provide further evidence of I-BET efficacy in an in vivo setting, highlighting further
potential clinical utility of BET inhibitors beyond the oncology arena.
0.8
0.7
0.6
*
0.5
.4
0.3
.2
0.1
.0
0
***
***
0
0
Vehicle Rapamycin I-BET762
mg/kg 3 mg/kg
I-BET762
10 mg/kg
3
ACS Paragon Plus Environment
27

Page 29 of 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 3. Effect of I-BET762 on ear swelling 24 h post-KLH challenge on Day 8. Results are
presented as mean ± SEM (n=10) of the difference between the KLH-injected ear and the NaCl
0.9 %-injected ear. Data were analysed by ANOVA followed by multiple comparisons using
Tukey’s post-hoc test (* p < 0.05, *** p < 0.001 versus vehicle).
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
Conclusions
In this report we have summarized our successful efforts in the optimization of the initial lead
GR90698X 7 culminating in the identification of the clinical compound I-BET762, 1.
Introduction and optimization of acetamide substituents in the 3-position of the benzodiazepine
core resulted in analogues stable under acid conditions and amenable to enantioselective
synthesis. SAR optimization produced I-BET762, 1, having the best overall potencies in the
BRD binding and cellular assays. This compound showed excellent preclinical development
6
properties and progressed into clinical development in oncology during 2012.
Experimental section
General. All commercial chemicals and solvents are reagent grade and were used without
further purification unless otherwise specified. All reactions except those in aqueous media were
carried out with the use of standard techniques for the exclusion of moisture. Reactions
performed under microwave irradiation utilized either a Biotage Initiator or CEM Discover
Microwave. Reactions were monitored by thin-layer chromatography on 0.2 mm silica gel plates
(ALUGRAM SIL G/UV254, Macherey-Nagel) and visualized with UV light. Final compounds
ACS Paragon Plus Environment
28