Structure-guided discovery of a selective mcl-1 inhibitor with cellular activity

Article abstract of DOI:10.1021/acs.jmedchem.9b00134

Myeloid cell leukemia 1 (Mcl-1), an antiapoptotic member of the Bcl-2 family of proteins, whose upregulation when observed in human cancers is associated with high tumor grade, poor survival, and resistance to chemotherapy, has emerged as an attractive target for cancer therapy. Here, we report the discovery of selective small molecule inhibitors of Mcl-1 that inhibit cellular activity. Fragment screening identified thienopyrimidine amino acids as promising but nonselective hits that were optimized using nuclear magnetic resonance and X-ray-derived structural information. The introduction of hindered rotation along a biaryl axis has conferred high selectivity to the compounds, and cellular activity was brought on scale by offsetting the negative charge of the anchoring carboxylate group. The obtained compounds described here exhibit nanomolar binding affinity and mechanism-based cellular efficacy, caspase induction, and growth inhibition. These early research efforts illustrate drug discovery optimization from thienopyrimidine hits to a lead compound, the chemical series leading to the identification of our more advanced compounds S63845 and S64315.

Full text of DOI:10.1021/acs.jmedchem.9b00134

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC
Article
The structure guided discovery of a
selective Mcl-1 inhibitor with cellular activity
Zoltan Szlavik, Levente Ondi, Marton Csekei, Attila Paczal, Zoltan B Szabo, Gabor
Radics, James B Murray, James Davidson, I-Jen Chen, Ben Davis, Roderick E Hubbard,
Christopher Pedder, Pawel Dokurno, Allan E. Surgenor, Julia Smith, Alan Robertson,
Gaetane Le Toumelin-Braizat, Nicolas Cauquil, Marion Zarka, Didier Demarles, Francoise
Perron-Sierra, Audrey Claperon, Frederic Colland, Olivier Geneste, and András Kotschy
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 24 Jun 2019
Downloaded from http://pubs.acs.org on June 25, 2019
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 34
Journal of Medicinal Chemistry
1
2
3
4
The structure guided discovery of a selective Mcl-1 inhibitor with cellular activity
5
6
7
8
9
Zoltan Szlávik,1 Levente Ondi,1† Márton Csékei,1 Attila Paczal,1 Zoltán B. Szabó,1 Gábor Radics,1‡
James Murray,2 James Davidson,2 Ijen Chen,2 Ben Davis,2 Roderick E. Hubbard,2 Christopher Pedder,2
Pawel Dokurno,2 Allan Surgenor,2 Julia Smith,2 Alan Robertson,2 Gaetane LeToumelin-Braizat,3
Nicolas Cauquil,3 Marion Zarka,3 Didier Demarles,4 Francoise Perron-Sierra,1, Audrey Claperon,3
Frederic Colland,3 Olivier Geneste,3 András Kotschy1,*
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
1Servier Research Institute of Medicinal Chemistry, Záhony u. 7., H-1031 Budapest, Hungary,
2Vernalis (R&D) Ltd., Granta Park, Cambridge CB21 6GB, UK
3Institute de Recherche Servier, 125 Chemin de Ronde, 78290 Croissy-sur-Seine, France
4Technologie Servier, 27 Rue Eugène Vignat, 45000 Orleans, France
Abstract:
Myeloid cell leukemia 1 (Mcl-1), an antiapoptotic member of the Bcl-2 family of proteins, whose
upregulation when observed in human cancers is associated with high tumor grade, poor survival,
and resistance to chemotherapy, has emerged as an attractive target for cancer therapy. Here we
report the discovery of selective small molecule inhibitors of Mcl-1 that inhibit cellular activity.
Fragment screening identified thienopyrimidine amino acids as a promising but non-selective hit
that were optimized using NMR and X-ray derived structural information. The introduction of
hindered rotation along a biaryl axis has conferred high selectivity to the compounds and cellular
activity was brought on scale by offsetting the negative charge of the anchoring carboxylate group.
The obtained compounds described here exhibit nanomolar binding affinity and mechanism-based
cellular efficacy, caspase induction and growth inhibition. These early research efforts illustrate drug
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 34
1
2
3
4
5
6
discovery optimisation from thienopyrimidine hits to a lead compound, the chemical series leading
to the identification of our more advanced compounds S63845 and S64315
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
Introduction
Apoptosis, an evolutionary highly conserved form of programmed cell death, is an essential process
for the elimination of no longer needed and dangerous cells.1 Evasion of apoptosis is recognised as a
critical element of the development as well as sustained expansion of tumours, and also underlies
resistance to diverse anti-cancer treatments.2 Mcl-1 is a member of the Bcl-2 family, critical
regulatory proteins of the mitochondrial apoptotic pathway, and is frequently upregulated in
cancer.3 Moreover increased expression of the MCL1 gene through transcriptional or post-
transcriptional mechanisms was observed as a downstream consequence of several key oncogenic
pathways.4 Mcl-1 is needed to sustain the growth of diverse tumours, including acute myeloid
leukaemia (AML)5, MYC-6 or BCR-ABL-driven pre-B/B lymphomas,7 certain breast cancers as well as
Non Small Cell Lung Carcinoma (NSCLC) derived cell lines that carry MCL1 gene amplifications.8
Certain compounds that broadly inhibit gene transcription or protein translation exert their cytotoxic
effects in tumour cells (at least in part) by downregulating MCL1.9
In the clinic, the highly promising activity of the Bcl-2-selective inhibitor venetoclax (Venclexta®),
which led to its approval in relapsed/refractory Chronic Lymphocytic Leukemia (CLL) patients with
17p deletion and in AML, has validated the use of drugs that directly activate apoptosis in cancer
therapy.10 Until recently only compounds showing weak cellular potency on Mcl-1 (high µmolar
range) and therefore useful only as in vitro chemical tools were available.11 Starting in late 2016 a
series of potent and selective Mcl-1 inhibitors were disclosed (Scheme 1) some of which have
recently also entered clinical development.12 The long standing interest in Mcl-1 as a target and the
late emergence of Mcl-1 targeting drug candidates suggests that drugging Mcl-1 is not trivial.
Supporting this analysis, the present manuscript describes our early research efforts to optimise
2
ACS Paragon Plus Environment

Page 3 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
thienopyrimidine hits13 to a lead compound on the way to our more advanced compounds S63845
and S64315 detailing the pitfalls we have encountered in this process.
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
F
F
F
N
N
O
N
O
N
O
N
H
N
O
Cl
S
Cl
H
O
O
S
CO2H
O
HN
O
N
N
N
N
N
O
O
S
N
O
S
O
F
AMG176
AZD5991
S63845
Figure 1 The structure of some recently disclosed potent and selective Mcl-1 inhibitors with reported
in vivo efficacy
Results
Fragment screening against Mcl-1 and Bcl-2 identified a number of carboxylic acid containing
compounds from which the thienopyrimidine derivative 1a emerged as a hit. This compound
showed comparable affinity against both targets (Ki of 50 µM and IC50 of 164 µM respectively) while
being moderately selective vs Bcl-xL (Table 1). The near neighbour screen of analogues (1b-1g)
convincingly validated this series as a starting point for hit expansion. Replacing the ethyl group by a
phenyl (1b) in the 5 position maintained the Mcl-1 affinity while increased selectivity against Bcl-2.
The same was also true for substituted phenyl analogues 1c and 1d. Interestingly the para
substituted analogue (1c) had decreased selectivity towards Bcl-xL. The m-hydroxyphenyl analogue
with a reversed amino acid stereochemistry (1e) showed also very similar affinity towards Mcl-1
while its affinity towards Bcl-2 increased considerably. Increasing the size of the 5 substituent (1f-g)
was also well tolerated both by Mcl-1 and Bcl-2 and the observed affinity values were also
corroborated by orthogonal techniques (Table 1).
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 34
1
2
3
4
Table 1. Mcl-1, Bcl-2, Bcl-xL inhibition of early thienopyrimidine hits (1a-g)
5
6
AA
7
8
9
HO
NH
R
O
1a-g
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
S
N
R
AA
Mcl-11
Bcl-21
Bcl-xL1
1a
1b
Et
S-Me
S-Me
50
74
NM
110
44
418
359
NM
216
SMe
1c
S-Me
1d
1e
1f
S-Me
R-Me
S-Me
R-Me
21
45
437
5.0
98
NM
60%@2.5mM
33
CF3
OH
F
F
130
192,3
O
O
1g
6.12
50%@1mM
1 Ki measured in Mcl-1, Bcl-2 or Bcl-xL FP assay in µM unless incomplete assay curve. NM – not
measurable at the concentrations used in the assay (2.5 mM or solubility limit)
2 For Kd measured by HSQC NMR assay see Supporting Information
3 For Kd measured by ITC assay see Supporting Information
4
ACS Paragon Plus Environment

Page 5 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Having validated the thienopyrimidines as a fragment hit series, in the absence of structural
guidance we embarked on the systematic variation of the core substituents. The analysis of the
binding data obtained on a small set of analogues revealed that modifications in the 2 position were
well tolerated but led to a substituent dependent selectivity profile. The smaller methylsulfonyl (2a)
derivative maintained affinity for Mcl-1 and also gained comparable affinity for Bcl-xL while showing
a minor drop in affinity for Bcl-2. On the other hand the pyridyl (2b) and sulfonamidophenyl (2c)
derivatives were more efficient inhibiting Bcl-2 than Bcl-xL (Table 2).
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
The variation of the amino acid substituent in the 4 position (3a-e) was also well tolerated. There
was no clear preference for any of the targets and the chirality of the side chain seemed to have only
minor influence on the affinity. Changing alanine to serine (3a) had no effect on binding to Mcl-1 but
modified selectivity towards the other targets. The phenylglycine analogue 3b brought affinity down
to the mid-micromolar range for all three targets. The influence of the chirality on the strength and
selectivity of the binding was explored with the phenylalanine derivatives 3c and 3d. Variation of the
absolute stereochemistry had little or no effect on the affinity or selectivity, both compounds
showing a similar profile as 3b. The saturated analogue 3e behaved very similarly to its parent
compound 3c.
For comparison we have also synthesized some analogues that bear the 3-hydroxyphenyl
substituent in position 5 (4a-d). The R- (4a) and S-phenylglcine (4b) analogues behaved similarly
showing a mild target dependent preference for one or the other stereoisomer. Comparing with the
more developed 3b we observed a similar selectivity profile but a considerable (ca. 10-fold)
improvement of affinity against all targets. The stereochemistry of the amino acid moiety had little
effect – as already seen for 3c-3d. The trends were slightly different for the R- (4c) and S-
phenylalanine (4d) analogues. Although both enantiomers showed similar affinities and selectivity
profile, in comparison with 3c and 3d we observed an increased selectivity towards Mcl-1.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 34
1
2
3
4
5
Table 2. Mcl-1, Bcl-2, Bcl-xL inhibition of thienopyrimidine derivatives 2-4
6
7
8
9
R2
OR3
HO
NH
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
O
N
2-4
R1
S
N
R1
R2
R3
Mcl-11
Bcl-21
Bcl-xL1
SO2Me
2a
2b
R-Me
4-fluorophenyl
44
51
26
R-Me
R-Me
4-fluorophenyl
4-fluorophenyl
8.4
50
4.1
1.3
18
39
N
2c
SO2NEt2
3a
3b
3c
3d
3e
4a
4b
4c
4d
H
H
H
H
H
H
H
H
H
rac-CH2OH 4-fluorophenyl
61
68
51
77
53
7.7
27
10
58
71%
13
87
15
rac-Ph
(R)-Bn
(S)-Bn
4-fluorophenyl
4-fluorophenyl
24
24
4-fluorophenyl
24
18
(R)-CH2Cy
(R)-Ph
4-fluorophenyl
8.1
17
H
H
H
H
470
269
63%
68%
319
107
75%
73%
(S)-Ph
(R)-Bn
(S)-Bn
1 Ki measured in Mcl-1, Bcl-2 or Bcl-xL FP assay in µM (or % inhibition at 1 mM if incomplete assay
curve)
6
ACS Paragon Plus Environment

Page 7 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In the next step we assessed the effect of 6-substituents on the affinity and selectivity of our
compounds (Table 3). The absence of a substituent (5a) or the introduction of a phenyl group (5b)
were both tolerated, although the latter compound was more potent and showed less selectivity
towards Bcl-2. Further introduction of a benzene moiety on the amino acid (5c) showed no
significant effect. Returning to the ethyl substituent in the 6-position and replacing the phenol in the
5-position by a 2-naphthyl moiety (5d) had no beneficial effect on the affinity or selectivity of the
molecule but we were able to obtain structural information about the binding of our compounds for
the first time using an NMR guided model.
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 3 Mcl-1, Bcl-2, Bcl-xL inhibition of thienopyrimidine derivatives 5a-d
R2
HO
R3
NH
O
N
R4
S
5a-d
N
R2
R3
R4
Mcl-11
592
Bcl-21
4%@2.5mM
49
Bcl-xL1
287
5a
5b
5c
R-Me 3-hydroxy-phenyl
H
R-Me 3-hydroxy-phenyl Ph
R-Bn 3-hydroxy-phenyl Ph
272
42
24
21
71%@1mM
5d
R-Me
2-naphthyl
Et
1102
29%@2.5mM
20%@2.5mM
1 Ki measured in Mcl-1, Bcl-2 or Bcl-xL FP assay in µM or percentage inhibition at top concentration if
incomplete assay curve.
2 For Kd measured by HSQC NMR assay see Supporting Information
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 34
1
2
3
4
First structural insight – NMR Guided Model established for 5d in Mcl-1
5
6
7
8
The NOE distance constraints suggested that compound 5d is in contact with Mcl-1 via the naphthyl
ring and one of the two methyls (from either the 2- or 6- position). The residues giving rise to the
NOEs are indicated in Figure 2a. In Figure 2b, the surface of Mcl-1 is coloured in yellow to show
where the NOEs occurred. This suggests that compound 5d binds to a location similar to which
Leu10 and Ile13 (red sticks) from the bound Bim peptide (red tube) occupy.
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
Analysis of the available BH3 public structures at the beginning of this work already suggested a
significant conformational change between a peptide-bound to a compound-bound BH3 protein. Yet
when we started to work on compound 5d binding model, there were only peptide-bound Mcl-1
structures. To address the possibility of protein flexibility, a protein conformational ensemble was
prepared for Mcl-1 to model compound 5d. When enumerating such ensemble of conformations,
we focused particularly on the residues involved in the NOEs. Compound 5d was then docked into all
enumerated protein conformations. The resulting docking models were visually inspected and the
best three docking models in terms of satisfying the observed NOEs were then subjected to induced-
fit docking protocol. This was to further explore the complementarity between the compound 5d
and hypothesized Mcl-1 conformation.14 And this led to a final model shown in Figure 2c, with the
ethyl and naphthyl ring in close contact with Mcl-1 and the acid likely pointing towards the solvent.
The proposed binding mode is in line with most of the observed SAR (e.g. 2-position, 6-position) but
the observed Mcl-1 selectivity for 4c and its analogues is difficult to interpret. Of note is that the
final NGM starting from peptide-bound Mcl-1 structures deviates from the X-ray structures obtained
at a later stage of the project despite efforts made to explore the protein flexibility. Nevertheless,
the NGM offers a number of hypotheses to be tested.
8
ACS Paragon Plus Environment

Page 9 of 34
Journal of Medicinal Chemistry
1
2
3
4
(a)
(b)
(c)
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
Figure 2. NMR guided model of 5d. The observed NOEs between Mcl-1 and 5d are summarized in
(a). The broken arrows suggest that either of the methyls could contribute to the observed NOEs. (b)
The NOEs highlighted in yellow surface are mapped to the Bim-bound Mcl-1 (2PQK) in grey surface.
The Bim is shown in red tube with Leu10 and Ile13 highlighted in red sticks. (c) The final binding
model of 5d is shown in with yellow surface indicating the residues giving rise to the NOEs.
With the structural guidance in hand we set out to explore the limits and flexibility of the
hydrophobic S2 pocket. Fixing the 6-position substituent as ethyl to allow for some flexibility within
the pocket and staying with alanine in the 4-position first we evaluated the 5-indolyl derivative 6a.
This compound showed good affinity towards both Mcl-1 and Bcl-2, which was also verified by
orthogonal techniques. The replacement of the linking nitrogen atom of the amino acid by oxygen
(7a) improved Mcl-1 binding by 6-8 fold. Since aryl ethers have a different rotational barrier than the
one from anilines, it would seem that the ether oxygen may offer a more preferable torsional profile
for Mcl-1 binding, its bioactive torsion well aligned with its energy minimum. Changing the polarity
and flexibility of the 5-substituent by replacing indolyl with 2-naphthyl (7b) or p-isopropylphenyl (7c)
was also well tolerated resulting only in a minor drop of affinity. Probing the flexibility of the system
we also prepared the o-benzyloxyphenyl derivative in the amino acid series (6b). In the reaction a
mixture of diastereoisomers was formed that were separable and we isolated 6b as the later eluting
diastereoisomer. This compound has also maintained its affinity towards Mcl-1 hinting that through
a conformational mobility Mcl-1 might accommodate bulky but flexible ortho-substituents on the
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 34
1
2
3
4
5
6
7
8
9
aromatic ring. 6b showed a high selectivity against the other targets therefore in the next set of
compounds we introduced the o-tolyl moiety in the 5-position. These molecules also possess
hindered rotation around the biaryl axis and their atropoisomers might be separated. We have
isolated and tested the pair of diastereomeric compounds 7d and 7e, which showed very similar
affinity towards Mcl-1. The relative stereochemistry of the atropoisomers was not determined. The
amino acid analogue 6c was tested as a 3:7 mixture of atropoisomers and showed a comparable
affinity as the hydroxyl acid analogues. All the o-tolyl analogues (7d-e,6c) inherited the
advantageous selectivity profile of 6b.
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
10
ACS Paragon Plus Environment

Page 11 of 34
Journal of Medicinal Chemistry
1
2
3
4
Table 4 Mcl-1, Bcl-2, Bcl-xL inhibition of thienopyrimidine derivatives 6-7
5
6
7
8
9
HO
R3
X
X=
X=O 7a-e
6a-c
NH
O
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
S
N
R3
X
Mcl-11
Bcl-21
Bcl-xL1
6a
7a
NH
O
222,3
282
67%@2.5mM
H
N
2.8
6.0
NM
NM
10%@0.33mM
57%@1mM
7b
7c
6b
O
O
8.1
24
NM
41%@1mM
BnO
NH
20%@2.5mM 27%@2.5mM
7d
7e
6c
O
O
18
15
NM
NM
25%@1mM
34%@1mM
4%@1mM
NH
100
70%@1mM
1 Ki measured in Mcl-1, Bcl-2 or Bcl-xL FP assay in µM unless incomplete assay curve. NM – not
measurable at the concentrations used in the assay (2.5 mM or solubility limit)
2 For Kd measured by HSQC NMR assay see Supporting Information
3 For Kd measured by ITC assay see Supporting Information
At this point we also obtained the X-ray structures of the 2-indolyl analogues both in the amino (6a)
and hydroxy (7a) acid series (Figure 3). In both structures (6a and 7a) the 2-indolyl moiety points
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 34
1
2
3
4
5
6
7
8
9
towards the solvent, with Met231 just ~4Å away from the side of the ring. The indolyl ring is ~56 (7a)
to 58 (6a) degrees rotated from the thienopyrimidine’s plane, fitting into the groove between
Ala227 and Met231. Analysis of this newly revealed binding mode suggested that major
improvements could be achieved by the fine tuning of the amino/hydroxyl acid side chain and the 5-
substituent. D-phenylalanine was already shown to have a beneficial effect both on the affinity and
selectivity (e.g. 4c). Combining this feature with the 4-indolyl moiety in the 5-position gave separable
diastereoisomers 8a and 8b and we observed submicromolar affinity for the first time (confirmed by
ITC) and a high selectivity towards Mcl-1. The atropoisomers were only mildly differentiated. The
hydroxyl acid analogues 9a and 9b behaved alike with 490 and 640 nanomolar IC50s respectively. The
X-ray structures of 8a and 8b revealed (Figure 3) that the pocket can accommodate both
atropoisomers similarly. All major elements (e.g. the phenyl side chain, the thienopyrimidine core,
the ethyl core substituent, and the indole moiety) occupy the same space in the two structures,
which explains the similar observed affinity of the diastereoisomers. Based on earlier observations
the o-tolyl analogues were expected to show a similar behaviour. The respective diastereomers were
synthesized and tested in both the amino acid (8c-d) and hydroxyl acid (9c-d) series. For both pairs
of compounds we observed a more pronounced difference of affinity between the atropoisomers
the more active ones registering around 1 µM. The X-ray structure of the more active stereoisomer
8d in the bound form (Figure 3) was coherent with the previous structures and it revealed that the
methyl group points towards the protein forming hydrophobic contacts with Phe228 and Phe270.
Such hydrophobic contacts could explain the affinity difference between atropoisomers as the other
atropoisomer (8c) would lose such favourable interaction by projecting the methyl towards the
solvent leading to weaker affinity. We have investigated the stability of the atropoisomers and found
no interconversion by 1H NMR on heating the solution of 8d in deuterated dimethylsulfoxide at 373
K for 30 mins. This finding ensured that the interconversion of 8c-8d or 9c-9d doesn’t bias our
biophysical measurements at ambient temperature.
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
12
ACS Paragon Plus Environment

Page 13 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
Table 5 Mcl-1, Bcl-2, Bcl-xL inhibition of thienopyrimidine derivatives showing atropoisomerism in
the 5-position (8-9)
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
HO
X
R
X=NH 8a-d
X=O 9a-d
O
N
S
N
R
X
Isomer3
Mcl-11
Bcl-21
Bcl-xL1
8a
NH
1
1.62
70%@1mM
135
H
N
8b
9a
NH
O
2
1
5.72
77%@1mM 76%@1mM
0.492
NM
NM
9b
8c
8d
9c
9d
O
2
1
2
1
2
0.642
58
11%@10µM
14%@1mM
NM
NH
NH
O
4%@1mM
2.42
3.4
46%@1mM 22%@1mM
NM
NM
O
0.89
5%@10µM
2%@10µM
1 Ki measured in Mcl-1, Bcl-2 or Bcl-xL FP assay in µM unless incomplete assay curve. NM – not
measurable at the concentrations used in the assay (10 µM or solubility limit)
2 For Kd measured by ITC assay see Supporting Information
3 arbitrary assignment of diastereomers based on the chromatographic elution order
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 34
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
Figure 3. Crystal structures of compounds 6a (PDB code 6QXJ), 7a (6QYK), 8a (6QZ5, 6QYL), 8b (6QZ6,
6QZ7), 8d (6QZB), 10d (6QZ8, 6QYN), 13 (6QYP), and 18a (6QYO) bound to the S2 pocket of Mcl-1.
Structure determination details can be found in the supplementary material.
The bound structure of 8d suggested that we might pick-up further interaction with the protein
backbone (Ala227) by introducing a chlorine or bromine next to the methyl group. The 4-position of
the same benzene ring also offers a vector for projecting substituents towards the solvent (i.e. to
improve compound properties). The introduction of bromine onto 9c and 9d resulted in the
atropoisomers 10a and 10b respectively. The stereoisomer 10a showed very similar affinity to its
parent 9c, which is understandable considering that the bromine and the methyl group are
projected towards the middle of the pocket. In contrary the affinity of the isomer 10b was improved
suggesting that the bromine atom might pick up some interaction with the protein. Using the more
compact chlorine instead of the bromine (10c) was also well tolerated. As expected this halogenated
compound maintained the selectivity towards other family members. The affinity of the amino acid
analogue 11a was in the same range both by FP and ITC assays.
To further improve the affinity for Mcl-1 we assessed two approaches based on the available
structural information. First we prepared the 2-methoxy-D-phenylalanine analogue 11b. As expected
this modification was well tolerated showing very similar affinity and selectivity as the parent
compound (11a). The other modification was the replacement of the ethyl moiety with a 2-propenyl
14
ACS Paragon Plus Environment

Page 15 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
group (11c). The resulting compound showed unfortunately no improvement but maintained more
or less the affinity of the ethyl analogue. Finally we have tested the applicability of the benzene ring
in position 5 as a starting point towards the solvent. The introduction of the phenolic function (10d)
was not only tolerated but led to a significantly improved affinity for the target that was also
validated by orthogonal techniques. We also determined the bound structure of 10d (Figure 3),
which confirmed that the phenolic hydroxyl group should be a suitable vector towards the solvent.
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
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 34
1
2
3
4
Table 6 Mcl-1, Bcl-2, Bcl-xL inhibition of the advanced thienopyrimidine derivatives 10-11.
5
6
R
Z
7
8
9
Cl
Y
HO
HO
NH
N
X
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
X=O: 10a-d
X=NH: 11a-b
11c
O
O
N
N
S
S
Y
Z
R
H
H
H
H
Isomer3
X
O
Mcl-11
5.3
Bcl-21
Bcl-xL1
NM
10a
Br
Br
Cl
Cl
Cl
Cl
H
H
H
H
H
H
1
2
NM
NM
36
10b
10c
11a
11b
11c
10d
O
0.332
0.81
NM
O
62%@0.2mM
48%@0.2mM
1
NH
NH
NH
O
0.512
0.262
19
29
Ome
H
2
38%@0.2mM 11%@0.2mM
5.4 35%@0.5mM
61%@0.2mM 29%@0.2mM
Cl OH
H
0.0512
1 Ki measured in Mcl-1, Bcl-2 or Bcl-xL FP assay in µM unless incomplete assay curve. NM – not
measurable at the concentrations used in the assay (10 µM or solubility limit)
2 For Kd measured by ITC assay see Supporting Information
3 arbitrary assignment of diastereomers based on the chromatographic elution order
With the affinity of our compounds for Mcl-1 reaching the low nanomolar range we set-up cellular
viability assays in the H929 multiple myeloma cell line, an MCL1-dependent cell line12 as single agent
or in combination with ABT-263, which sensitizes H929 cells through the inhibition of other
important anti apoptotic proteins (Bcl-2, Bcl-xL, Bcl-w). To assess and mitigate the potential limiting
effect of plasma protein binding, some of the cellular assays were also run both at standard (10%)
and lower (0.1%) serum concentrations. Our compounds (e.g. 10d, 11b) showed no cellular activity,
16
ACS Paragon Plus Environment

Page 17 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
which we linked to poor anticipated cellular penetrance. In order to improve this property we
explored the introduction of a basic group onto our molecules. This modification was expected to
offset the negative charge of the carboxylate group under physiological conditions. On analysis of
the bound structure of 10d (Figure 4), we concluded that this additional moiety could be introduced
either into the ortho-position of the benzene ring on the amino/hydroxyl acid or on the tolyl moiety
in the 5-position.
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
Our selected permeabiliser, the methylpiperazine moiety was first connected to 11b using a two
carbon linker (12). In spite of a significant loss of affinity (0.36 µM cf. 0.024 µM by ITC) 12 registered
in our viability assays. Its IC50 of 20.3 µM under standard conditions was slightly improved both by
the addition of ABT-263 (12 µM) and by the decrease of the serum concentration in the experiment
to 0.1% (9.8 µM; single agent). Using the tolyl moiety as the anchoring point and connecting the
piperazine to its 5-position either directly (13) or through a methylene linker (14) resulted in the
same beneficial effect. We have obtained the X-ray structure of 13 bound to Mcl-1 (Figure 3), which
showed that the piperazine moiety is nicely accommodated in the solvent exposed part of the
binding pocket and it has not changed the way the ligand is bound. Both 13 and 14 registered in the
cellular assays as single agent under standard conditions (13.3 µM and 15.5 µM at 10% serum) and
this effect was attenuated in combination with ABT-263 (4.2 µM and 5.4 µM respectively). Using the
phenolic hydroxyl group of 10d as an anchor point and linking it through an acetyl moiety (15) was
also very well tolerated. The observed affinity was again very similar to the parent phenol while the
compound registered in the cellular assays in the low micromolar range. The tolerance of the
substitution of both the 4- and 5-positions of the tolyl moiety suggested that cyclization of the two
positions should also be tolerated. Indeed the benzofurane analogue 16 showed comparable
characteristics to its “parents” 14 and 15 both affinity wise and considering its cellular activity.
Finally we have incorporated a simple ethylene linker to connect the N-methylpiperazine moiety to
the phenolic oxygen. Of this compound both the amino acid (17) and the hydroxyl acid (18)
analogues were prepared.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 34
1
2
3
4
5
6
7
8
9
To better understand the behaviour of these compounds and their biological effect, we isolated and
tested both atropoisomers (17a-b, 18a-b). Comparing the amino (17a) and hydroxyl (18a) acids the
latter showed both superior affinity towards Mcl-1 and more efficient cell killing in the viability
assay. As expected the atropoisomeric compounds 17b and 18b showed a significantly decreased
affinity for Mcl-1, which made them good controls in the following pharmacological studies. The
absolute configuration of 18a was also confirmed by determination of the structure of its Mcl-1
bound complex (Figure 3), which showed that the methyl group and the chlorine atom interact with
the protein surface while the piperazine unit is projected towards the solvent as expected. This
proved that the para position (Z in Table 6) of the 5-phenyl moiety is able to modulate physical
properties and cellular activities of the series, while maintaining the same mode of binding.
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
18
ACS Paragon Plus Environment

Page 19 of 34
Journal of Medicinal Chemistry
1
2
3
4
Table 7. Mcl-1 inhibition and cell killing of thienopyrimidine derivatives (12-18) in the presence and
5
6
absence of ABT-263
7
8
9
MTT w ABT-
Structure
Mcl-11
MTT IC50
3
2
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
263 IC50
N
O
N
Cl
12
2.04
12.2
20.3/9.8
HO
NH
N
O
N
S
N
OH
N
Cl
13
0.0464
4.2
13.6/4.7
HO
O
N
O
N
S
N
N
OH
14
0.0324
5.4
15.5/NA
Cl
HO
O
N
O
N
S
N
O
N
O
Cl
0.0214
6.9
17.1/NA
HO
15
O
O
N
S
N
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 34
1
2
3
4
5
N
6
7
O
8
9
16
0.0764
10
20.4/3.8
11.6/1.7
Cl
HO
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
O
N
O
N
S
N
O
N
17a
0.224
2.5
Cl
H O
NH
N
O
N
17b
18a
170
--
>30µM/>30µM
5.6/1.0
S
0.0194
1.1
N
O
N
Cl
HO
O
N
O
18b
25
23.4
27.2/NA
N
S
1 Ki measured in Mcl-1 FP assay in µM
2 IC50 measured in cell viability assay in µM in H929 cells following 48h incubation in the presence of
1 µM ABT-263.
3 IC50 measured in cell viability assay in µM in H929 cells following 48h incubation in the presence of
10%/0.1% serum.
4 For Kd measured by ITC and SPR assay see Supporting Information
The finding that 18a inhibits Mcl-1 and induces cell death when used as single agent in H929 cancer
cells at single digit micromolar level prompted us to probe its behaviour in more detail. To ensure
that we have a real Mcl-1 binder we ran two different binding assays with Mcl-1. In the ITC
experiment we measured a Kd of 30 nM, while in an SPR experiment with immobilized Mcl-1 we
observed a Kd of 10 nM with a dissociation half-life of 17.5 second. Both of these data align well with
the Ki of 19 nM measured in the primary FP assay.
20
ACS Paragon Plus Environment

Page 21 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The next experiment aimed at proving that 18a acts in cells through displacing selectively a relevant
proapoptotic BH3 domain containing protein from its Mcl-1 complex. To this end, HeLa cells
overexpressing Flag-Mcl-1 or Flag-Bcl-xL were treated with different doses of 18a and 18b. Following
immunoprecipitation using anti-Flag antibody, the endogenous Bak protein complexed with either
Mcl-1 or Bcl-xL was monitored (Figure 4a). Interestingly, the Mcl-1/Bak complex was disrupted
following treatment with compound 18a, as evidenced by the dose-dependent decrease of
endogenous Bak co-immunoprecipitated with Flag-Mcl-1.In the same conditions, no effect was
observed on the Bcl-xL/Bak complex. As negative control, treatment with its corresponding less
active atropoisomeric compound 18b did not affect the Mcl-1-Bak complex.
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
Selective Mcl-1 inhibitors were reported to stabilize the level of endogenous Mcl-1 protein in a dose-
dependent manner.11a,12a, 12b To monitor this target hitting, colon carcinoma HCT116 cell line (not
sensitive to Mcl-1 inhibition12a) was treated with different doses of Mcl-1 inhibitors 17a and 18a and
their corresponding less active diastereoisomers 17b and 18b. Significant dose-dependent increase
of the endogenous Mcl-1 protein was observed following treatment with either compound 17a or
18a (Figure 4b). Interestingly, this assay is clearly more sensitive than the viability assay since the
two active compounds exhibited activity at doses as low as 0.3 µM. The lack of concomitant PARP-
cleavage confirmed the absence of apoptosis induction in this cell line resistant to Mcl-1 inhibition.
Importantly, the less active atropoisomers (17b, 18b) exhibited no effect on the Mcl-1 protein level
thus further supporting Mcl-1 target hitting with active compounds (17a and 18a) in cells.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 34
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
Figure 4. a) Evaluation of Mcl-1 inhibitors on Mcl-1/Bak and Bcl-xL/Bak complexes by co-
immunoprecipitation; Arrow indicates Flag-MCL-1; b) Dose-dependent Mcl-1 stabilisation in HCT116
cells following treatment with 17a and 18a compounds and their less active diastereoisomers (17b,
18b).
Once Mcl-1 target hitting was demonstrated with compounds 17a and 18a, we next asked whether
these Mcl-1 inhibitors could induce apoptosis in a Mcl-1 sensitive cell line. To this end, H929 multiple
myeloma cell line was treated for 6h with different doses of 17a, 17b, 18a, and 18b in the presence
22
ACS Paragon Plus Environment

Page 23 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
or absence of ABT-263, which was expected to further sensitize cells to Mcl-1 inhibition. Two
different but commonly accepted apoptosistic cellular readouts were used: PARP cleavage and
cleaved Caspase-3. As shown in Figure 5, 17a and 18a induced both PARP and Caspase-3 cleavage in
this cell line in a dose-dependent manner. This effect was amplified by the addition of the Bcl-2/Bcl-
xL inhibitor ABT-263 and completely abrogated by the caspase inhibitor Q-VD-OPh (QVD).
Importantly, the less active diastereomers showed no apoptosis induction, in agreement with our
previous findings. Altogether, these data suggest that 17a and 18a are selective Mcl-1 inhibitors
capable of demonstrating on-target cell killing in H929 cancer line through activation of the
apoptotic pathway.
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 5. Dose-dependent apoptosis induction in H929 cells following treatment with Mcl-1
inhibitors 17a and 18a and their less active diastereoisomers (17b, 18b) in the presence or absence
of ABT-263 and caspase inhibitor QVD.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 34
1
2
3
4
5
6
7
8
Based on their affinity and cellular activity, ADME properties of 17a and 18a were further
characterized. Their hepatic microsomal clearance in mice or rat was good to acceptable with a clear
superiority of 18a (Table 8, 30 and 11 ml/min/kg vs 72 and 23 ml/min/kg respectively). In the
presence of human microsomes 18a was also less metabolised than 17a (10 vs 18 ml/min/kg)
although both values were inferior compared to the rodent species. As it is typical for PPI inhibitors15
both compounds have a very low free fraction in plasma (18a: 0.2% in mice and 0.5% in human vs
0.2% and 0.2% for 17a respectively). The stability of 17a and 18a in the presence of hepatocytes
without added plasma was low for human (18 and 17 ml/min/kg) while acceptable to good for mice
(75 and 9 ml/min/kg) and rat (42 and 20 ml/min/kg) showing again a superiority of the hydroxyl acid.
As one might expect from the high PPB values the metabolic clearance of both 17a and 18a became
good in hepatocytes in the presence of added plasma (mice: 3 and 0 ml/min/kg, rat: 24 and 0
ml/min/kg, human: 0 and 5 ml/min/kg). The predicted human intestinal absorption of 17a and 18a in
the Caco-2 model were 59% and 29% with mass recoveries at 82% and 77%, respectively.
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
human
mice
rat
CL pred (ml/min/kg) FU(%) Fabs(%) CL pred (ml/min/kg) FU(%) CL pred (ml/min/kg)
MIC
18
HEP
MIC
72
HEP
75/3#
9/0#
MIC
23
HEP
42/24#
20/0#
17a
18a
18/0#
17/5#
0.2
0.5
59
29
0.2
0.2
10
30
11
Table 8. In vitro ADME parameters of 17a and 18a (MIC-microsomes, HEP-hepatocytes). #-
experiment run with added plasma
To test the predictive power of the in vitro ADME data, in vivo PK studies were also run in mice. The
compounds (17a, 18a) were dosed 1 mpk i.v. and 3 mpk p.o. On i.v. administration, both compounds
showed a similar profile (Table 9) while 18a was more persistent than 17a as manifested in the
24
ACS Paragon Plus Environment

Page 25 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
different PK parameters. This finding is in line with the difference in metabolic stability observed in
vitro, although the clearance of 17a was higher than one would estimate based on hepatocyte data.
The in vitro data suggested that the difference between 17a and 18a exposure would be levelled on
oral dosing, which was indeed the case. The calculated oral bioavailabilities were 16% and 9%
respectively which indicated limited absorption as suggested by in vitro Caco-2 results.
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
i.v. 1mpk
AUCt
p.o. 3mpk
t1/2 AUCt
F%
p.o./
i.v.
t1/2
h
Cl plasma
Vd
blood/
Cmax
mL/min/kg
ng/mL.h
L/kg
plasma ng/mL
ratio
h
ng/mL.h
17a
18a
0.65
3.7
33
509
0.9
1.0
0.7
0.6
64
3.8
1.7
168
781
16%
9%
5.7
2863
199
Table 9. in vivo PK parameters of 17a and 18a in mice.
To assess the drug-drug interaction and off-target profile of our leads, potential inhibitory effect of
these compounds was evaluated on 5 different human CYP450 enzymes (3A4, 2D6, 1A2, 2C9, 2C19).
17a inhibited only CYP 3A4 with an IC50 of 5.2 µM while the IC50 of 17a or 18a against the remaining
enzymes was not reached until 20 µM. 18a was also submitted to an hERG patch-clamp assay and
showed a dose-dependent inhibition reaching 28% at 10 µM concentration. Neither a safety
receptors profile nor in silico genotoxicity assessment of 18a showed any alerts. Based on these
data, 18a appeared as a promising lead compound ready for more extended optimization.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 34
1
2
3
4
Discussion and Conclusions
5
6
7
8
An NMR-based fragment screen identified thieno[2,3-d]pyrimidines bearing an amino acid in the 4-
position and a small alkyl substituent in the 6-position as an interesting, non-selective hit for Mcl-1.
A systematic variation of the available positions and use of orthogonal assays confirmed the hit
series and allowed the initiation of a fragment growing program. In the early stages, varying
selectivities were observed and typically Mcl-1 and Bcl-2 were inhibited alike. Structural guidance
was provided by NMR-guided model building and later by X-ray crystallography. It is interesting to
note that the presence of different binding modes was observed at this stage of the development. A
significant breakthrough came with the introduction of such aromatic substituents into the 5-
position of the thienopyrimidine core that showed restricted rotation due to the presence of an
ortho-substituent. These molecules were highly selective for Mcl-1 and affinities shifted to the sub-
micromolar range. Structure-guided fine-tuning of the inhibitors and the introduction of a basic
nitrogen into the solvent exposed part of the molecule to offset the charge of the carboxylic acid
under physiological conditions led to potent and selective inhibitors which induced cell killing.
Cellular efficacy was observed in correlation with the expected mechanism-based events such as
Mcl-1 stabilization, caspase induction and PARP cleavage. The combined data of cellular experiments
suggest that to induce apoptosis in Mcl-1 dependent cell lines significant higher doses than the Kd of
the inhibitor are required, which is probably linked to the need of replacing most of the BH3-only
proapoptotic proteins from their complex with Mcl-1. The pharmacokinetic properties of the lead
compounds were determined as well as any potential DDI liability. Since favourable results were
obtained in all assays, these lead compounds were further developed leading to the discovery of
anti-Mcl-1 clinical candidate.
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
Experimental section
General
26
ACS Paragon Plus Environment

Page 27 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
All obtained products had an LC purity above 96% that was corroborated by their 1H NMR spectrum
unless specifically mentioned otherwise. All synthetic experimental details including the
characterisation of the compounds are described in the Supporting Information.
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
Pharmacology material and method
Compounds. ABT-263 was purchased from Selleck-chem and QVD-OPh from Sigma.
Cell culture. NCI-H929, Hela andHCT-116, cells were cultured in RPMI 1640 medium supplemented
with 10% heat inactivated FBS, 2 mM L-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin, and
10 mM Hepes, pH = 7.4 at 37 °C, in 5% CO2/95% air. Cells were grown at 37 °C in a humidified
atmosphere with 5% CO2. All of these cell lines were purchased from the ATCC.
MTT cell viability assay. Cell viability was measured using MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide) colorimetric assay. Cells cultures either in 0.1% or 10% serum were
seeded in 96-well microplates at a density to maintain control (untreated) cells in exponential phase
of growth during the entire experiment. Cells were incubated with compounds for 48 h followed by
incubation with 1 mg/mL MTT for 4 h at 37 °C. Lysis buffer (20% SDS) was added and absorbance
was measured at 540 nm 18 h later. All experiments were repeated at least 2 times in triplicates. The
percentage of viable cells was calculated and averaged for each well: % growth = (O.D. treated
cells/O.D. control cells) x 100, and the IC50, concentration reducing by 50% the optical density, was
calculated by a linear regression performed on the linear zone of the dose-response curve.
Co-immunoprecipitation
HeLa cells were transiently transfected, using Effecten reagent (Qiagen), with 3xFlag-tagged BCL-XL
or MCL-1 expression vectors (p3xFlag- CMV10, Sigma). 24 h later, transfected Hela cells were treated
with 18a or 18b during 2 h and harvested in lysis buffer (10 mM Hepes pH 7.5, 150 mM KCl, 5 mM
MgCl , 1 mM EDTA, 0.4% TritonX100), protease and phosphatase inhibitors cocktails (Calbiochem
2
539134 and 524625). HeLa cleared lysates were then subjected to immunoprecipitation with anti-
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 34
1
2
3
4
5
6
Flag M2 agarose beads (Sigma). The immunoprecipitates and inputs were analyzed by immunoblot
using BAK antibody (BD 556396) or Flag M2 (Sigma).
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
Mcl-1 stabilisation. HCT-116 were incubated for 16h with Mcl-1 inhibitors. Total protein extracts of
HCT-116 cells were generated in lysis buffer (20 mM Tris-HCl, pH 7.4, 135 mM NaCl, 1.5 mM MgCl2, 1
mM EDTA, 10% glycerol) containing 1% Triton X-100 and complete protease inhibitors (Roche).
Protein extracts of the other cell lines were generated in lysis buffer containing 10 mM Hepes pH
7.4, 142.5 mM KCl, 5 mM MgCl2, 1 mM EDTA, 1% NP40, protease and phosphatase inhibitors
cocktails (Calbiochem). Protein content was quantified using the Bradford assay (Bio-Rad). Lysates
were diluted with LDS sample buffer (Invitrogen) at a 3:1 ratio and denatured at 95 C for 7-10 min.
30 g of protein extracts were separated by SDS:PAGE (NuPAGE 10% Bis Tris gels) and proteins
transferred onto nitrocellulose membranes. The membranes were blocked in 5% skimmed milk in
PBS and 0.1% Tween20 (blocking buffer) before incubation with antibodies. Commercially available
antibodies were used: rabbit polyclonal antibodies against Mcl-1 (Santa Cruz S-19, sc-819), PARP
(Cell Signaling 9542), and mouse monoclonal antibodies against Actin (Millipore MAB1501R; used as
a loading control).
Immunodetection of cleaved PARP by MesoScale discovery assay. H929 cells were treated with
QVD and the indicated compounds in addition to ABT-263 1µM for 6 h and harvested in lysis buffer
(10 mM Hepes, pH 7.4, 142.5 mM KCl, 5 mM MgCl2, 1 mM EDTA, 1% NP40, protease and
phosphatase inhibitors cocktails (Calbiochem). Cleared lysates (5 µg protein) were prepared for
immunodetection of cleaved PARP or cleaved caspase 3 (markers of apoptosis) by using the MSD
Apoptosis Panel Whole Cell Lysate kit (MSD) in 96-well plates according to manufacturer’s
instructions, and analysed on the Sector Image 2400.
ASSOCIATED CONTENT
28
ACS Paragon Plus Environment

Page 29 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Supporting Information. Supplementary information with description of chemical synthesis,
analysis, co-immunoprecipitation study, structural determination details and the NMR guided model
as well Molecular Formula Strings.
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
Accession Codes. The X-ray structures mentioned in this paper have been deposited in the PDB with
the following codes: 6QXJ, 6QYK, 6QZ5, 6QYL, 6QZ6, 6QZ7, 6QZB, 6QZ8, 6QYN, 6QYP, 6QYO.
AUTHOR INFORMATION
Corresponding Author
*Phone: +36 (1) 881-2000. Fax: +36 (1) 881 2011. E-mail: andras.kotschy@servier.com.
ORCID Andras Kotschy: 0000-0002-7675-3864
Present Addresses
†Levente Ondi: XiMo Hungary Ltd., Záhony u. 7., H-1031 Budapest, Hungary
‡Gábor Radics: Hunyadi János u. 34, H-2030 Érd, Hungary
Author contributions
The manuscript was written through contributions of all authors. All authors have given approval to
the final version of the manuscript.
ORCID for other authors
ORCID Roderick Hubbard: 0000-0002-8233-7461
ORCID Marton Csekei 0000-0002-5781-1096
ORCID Gabor Radics 0000-0003-4954-3025
ORCID Zoltan B Szabo 0000-0001-7557-0305
ORCID Zoltan Szlavik 0000-0002-9385-806X
■ ACKNOWLEDGMENTS
The authors thank co-workers at the Analytical Division of the Servier Research Institute of Medicinal
Chemistry for providing the detailed chemical analysis of the compounds.
■ ABBREVIATIONS USED
29
ACS Paragon Plus Environment