Discovery of Novel 3-Quinoline Carboxamides as Potent, Selective, and Orally Bioavailable Inhibitors of Ataxia Telangiectasia Mutated (ATM) Kinase

Article abstract of DOI:10.1021/acs.jmedchem.6b00519

A novel series of 3-quinoline carboxamides has been discovered and optimized as selective inhibitors of the ataxia telangiectasia mutated (ATM) kinase. From a modestly potent HTS hit (4), we identified molecules such as 6-[6-(methoxymethyl)-3-pyridinyl]-4-{[(1R)-1-(tetrahydro-2H-pyran-4-yl)ethyl]amino}-3-quinolinecarboxamide (72) and 7-fluoro-6-[6-(methoxymethyl)pyridin-3-yl]-4-{[(1S)-1-(1-methyl-1H-pyrazol-3-yl)ethyl]amino}quinoline-3-carboxamide (74) as potent and highly selective ATM inhibitors with overall ADME properties suitable for oral administration. 72 and 74 constitute excellent oral tools to probe ATM inhibition in vivo. Efficacy in combination with the DSB-inducing agent irinotecan was observed in a disease relevant model.

Full text of DOI:10.1021/acs.jmedchem.6b00519

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Article
Discovery of Novel 3-Quinoline Carboxamides as Potent, Selective and
Orally Bioavailable Inhibitors of Ataxia Telangiectasia Mutated (ATM) Kinase
Sébastien L. Degorce, Bernard Barlaam, Elaine Cadogan, Allan Paul Dishington, Richard
Ducray, Steven C. Glossop, Lorraine A. Hassall, Franck Lach, Alan Lau, Thomas
M. McGuire, Thorsten Nowak, Gilles Ouvry, Kurt G. Pike, and Andrew G. Thomason
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00519 • Publication Date (Web): 03 Jun 2016
Downloaded from http://pubs.acs.org on June 4, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Novel 3ꢀQuinoline Carboxamides as
Potent, Selective and Orally Bioavailable Inhibitors
of Ataxia Telangiectasia Mutated (ATM) Kinase
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
Sébastien L. Degorce,*,†,‡ Bernard Barlaam,† Elaine Cadogan,† Allan Dishington,† Richard
Ducray,‡ Steven C. Glossop,† Lorraine A. Hassall,† Franck Lach,‡ Alan Lau,† Thomas M.
McGuire,† Thorsten Nowak,† Gilles Ouvry,‡ Kurt G. Pike,† and Andrew G. Thomason.†
†
Oncology Innovative Medicines Unit, AstraZeneca, Mereside, Alderley Park, Macclesfield,
‡
Cheshire SK10 4TG, United Kingdom; Oncology Innovative Medicines Unit, AstraZeneca,
Centre de Recherches, Z.I. la Pompelle, BP1050, 51689 Reims Cedex 2, France.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 42
1
2
3
4
ABSTRACT
5
6
7
8
9
A novel series of 3ꢀquinoline carboxamides has been discovered and optimized as selective
inhibitors of the Ataxia Telangiectasia Mutated (ATM) Kinase. From a modestly potent HTS hit
(4), we identified molecules such as 6ꢀ[6ꢀ(methoxymethyl)ꢀ3ꢀpyridinyl]ꢀ4ꢀ{[(1R)ꢀ1ꢀ(tetrahydroꢀ
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
2Hꢀpyranꢀ4ꢀyl)ethyl]amino}ꢀ3ꢀquinolinecarboxamide
(72)
and
7ꢀfluoroꢀ6ꢀ[6ꢀ
(methoxymethyl)pyridinꢀ3ꢀyl]ꢀ4ꢀ{[(1S)ꢀ1ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀyl)ethyl]amino}quinolineꢀ3ꢀ
carboxamide (74) as potent and highly selective ATM inhibitors with overall ADME properties
suitable for oral administration. 72 and 74 constitute excellent oral tools to probe ATM inhibition
in vivo. Efficacy in combination with the DSBꢀinducing agent irinotecan was observed in a
disease relevant model.
ACS Paragon Plus Environment
2

Page 3 of 42
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
The Ataxia Telangiectasia Mutated (ATM) kinase1 is an atypical serine/threonine protein
kinase involved in DNA damage responses (DDR).2 It belongs to the soꢀcalled PIKK
(PhosphatidylInositol 3’ꢀKinase [PI3K]ꢀrelated kinase) family, together with DNAꢀdependent
Protein Kinase (DNAPK), Ataxia Telangiectasia mutated and RAD3ꢀrelated (ATR) kinase,
themselves also key DDR targets, and mammalian Target Of Rapamycin (mTOR). ATM detects
DNA double strand breaks (DSBs),3 and signals activation of downstream DNA repair and cell
cycle checkpoint. As DSBs can be caused by ionizing radiation (IR) or chemically induced DNA
damage (e.g. the DNA topoisomerase I inhibitor irinotecan4), the inhibition of ATM in
combination with clinically induced DNA damage in tumour cells could bring important benefits
to cancer patients by potentiating the effects of such agents or techniques.5, 6
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
Few inhibitors of ATM have been previously reported, and their use has been limited by their
weak potency and/or lack of PIKKꢀselectivity, sometimes combined with unsuitable
pharmacokinetic properties. Amongst those, 1 (KUꢀ55933)7 and its improved version 2 (KUꢀ
60019)8 from Kudos Pharmaceuticals (now part of AstraZeneca) and the structurally distinct
quinazoline 3 (CPꢀ466722)9 from Pfizer are probably the most studied. Pyranꢀ4ꢀone 1 is a potent
and reasonably selective inhibitor of ATM with demonstrated activity in cells, but suffers from
poor solubility, high turnover in hepatocytes, and therefore could not be used as an in vivo tool
orally. It was later optimized to 2, a roughly 10ꢀfold more potent ATM inhibitor with higher
solubility and reduced hERG activity, but in our hands remained a short halfꢀlife molecule
despite improved metabolic stability (Table 1). After oral dosing, only low exposure in mice
could be achieved, insufficient to achieve durable ATM inhibition. Quinazoline 3 was, in our
hands, less potent both in enzyme and cell assays than 2, and with similar selectivity against
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 42
1
2
3
4
5
6
7
8
9
PI3K and PIKK kinases; however, despite lower lipophilicity 3 also suffered from moderate to
high turnover in rodent hepatocytes, which made it unsuitable as an in vivo tool. Other panꢀPI3K
and/or panꢀPIKK inhibitors showing good pharmacokinetic properties have also been reported to
inhibit ATM.10 Unfortunately, these could not be used as ATM in vivo probes because of their
lack of specificity. Our ambition was therefore to identify a novel, potent and selective ATM
inhibitor with good oral bioavailability to evaluate the potential of more specific ATM inhibition
in vivo.
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 1. Selected examples of reported ATM inhibitors.
ACS Paragon Plus Environment
4

Page 5 of 42
Journal of Medicinal Chemistry
1
2
3
4
SYNTHESIS
5
6
7
8
Compounds reported herein were synthesized as shown in Schemes 1–4. 3ꢀQuinoline
carboxamides substituted with an aryl group at C6 and an amine at C4 were generally made
using the versatile 6ꢀbromoꢀ4ꢀchloroquinolineꢀ3ꢀcarboxamide 11 via Suzuki crossꢀcoupling with
aryl boronic acids/esters followed by nucleophilic aromatic substitution (SNAr) with primary
amines, or by reversing these steps. In some cases where the aryl boronic acids/esters were either
unavailable or unreactive, the Suzuki crossꢀcoupling was reversed and 6ꢀ(4,4,5,5ꢀtetramethylꢀ
1,3,2ꢀdioxaborolanꢀ2ꢀyl)ꢀ3ꢀquinolinecarboxamides (e.g. 30) and aryl halides were used instead,
as described in Scheme 1.
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
6ꢀ(3ꢀPyridinyl)ꢀ3ꢀquinolinecarboxamides 13-23 were obtained via Suzuki crossꢀcoupling first
to give key intermediate 4ꢀchloroꢀ6ꢀ(3ꢀpyridinyl)ꢀquinolineꢀ3ꢀcarboxamide 12, followed by
SNAr with the appropriate primary amine or thiol. Synthesis of the 6ꢀpyrazole analogue 25 was
achieved in a similar manner, whilst 27-29 were obtained via SNAr first to give key intermediate
6ꢀbromoꢀ4ꢀ[(2ꢀethylbutyl)amino]ꢀ3ꢀquinolinecarboxamide 26, followed by Suzuki crossꢀ
coupling. Compounds 31-37 were made from 4ꢀ[(2ꢀethylbutyl)amino]ꢀ6ꢀ(4,4,5,5ꢀtetramethylꢀ
1,3,2ꢀdioxaborolanꢀ2ꢀyl)ꢀ3ꢀquinolinecarboxamide
30
and
the
corresponding
aryl
bromides/chlorides. 4ꢀ[(2ꢀEthylbutyl)amino]ꢀ6ꢀ(5ꢀpyrimidinyl)ꢀ3ꢀquinolinecarboxamide 33 was
also obtained in this way using (2ꢀchloroꢀ5ꢀpyrimidinyl)boronic acid in the crossꢀcoupling step,
with subsequent deꢀhalogenation by hydrogenation. Likewise, crossꢀcoupling with (5ꢀbromoꢀ2ꢀ
pyridinyl)methanol
provided
intermediate
6ꢀ[6ꢀ(hydroxymethyl)ꢀ3ꢀpyridinyl]ꢀ3ꢀ
quinolinecarboxamide 35. After bromination of the alcohol, the methoxy analogue 36 and the
methylsulfonyl analogue 37 were obtained by nucleophilic displacement with sodium methoxide
or sodium methanesulfinate.
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 42
1
2
3
4
5
6
7
8
9
Oꢀlinked 3ꢀquinolinecarboxamide 40 was obtained using ethyl 6ꢀbromoꢀ4ꢀchloroquinolineꢀ3ꢀ
carboxylate 38 via displacement of the 4ꢀchloro with 2ꢀethylbutyl alcohol with NaH in THF
followed by treatment of the ester with aqueous ammonia (Scheme 2). 8ꢀMethyl quinoline 43
was also obtained using the corresponding commercially available 3ꢀcarboxylate ester 41 which
was saponified and reacted with thionylchloride and then aqueous ammonia to give the
corresponding 4ꢀchloroꢀ6ꢀbromoꢀ3ꢀcarboxamide intermediate 42. Similar chemistry as shown in
Scheme 1 was subsequently applied to achieve the desired substitution at C4 and C6.
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
Further substitution on the quinoline core was completed by constructing the core using the
appropriate anilines and acrylates to afford the desired key 4ꢀchloroꢀ6ꢀbromo intermediates
(Scheme 3). In the case of analogues 49-51, the 3ꢀcyano quinolones 46-48 were formed by
condensation of the appropriate aniline and either ethylꢀ2ꢀcyanoꢀ3ꢀethoxybutꢀ2ꢀenoate (46) or
ethylꢀ2ꢀcyanoꢀ3ꢀethoxyacrylate (47-48). Conversion of the 3ꢀcyano into the desired 3ꢀ
carboxamide was performed using acetaldoxime and palladium acetate. Analogues 58-60 were
obtained in a similar way via condensation of the appropriate aniline and diethyl
(ethoxymethylene)malonate, followed by cyclisation in diphenylether at high temperature. The
3ꢀesters were then treated as described above to afford the 4ꢀchloroꢀ6ꢀbromoꢀ3ꢀquinoline
carboxamides 55-57 which were substituted at C4 and C6 in the usual way.
Combinations 72-76 shown in Scheme 4 were performed in a similar manner as described
above, using the handcrafted 2ꢀ(methoxymethyl)ꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)pyridine 63 (obtained in two steps from (5ꢀbromoꢀ2ꢀpyridinyl)methanol 61) or 2ꢀfluoroꢀ6ꢀ
(methoxymethyl)ꢀ3ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine 64 (also obtained in
two steps from 3ꢀbromoꢀ6ꢀ(bromomethyl)ꢀ2ꢀfluoroꢀpyridine 62). Likewise, 1,3,4ꢀtrisubstituted
pyrazoles 77ꢀ78 were synthesized in a similar way using the corresponding amines:
ACS Paragon Plus Environment
6

Page 7 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
fluoropyrazole 67 was obtained via fluorination of tertꢀbutyl [(1S)ꢀ1ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀ
yl)ethyl]carbamate 65 with Selectfluor; methylpyrazole 68 was obtained in a three step
procedure involving reductive amination of 1ꢀ(1,4ꢀdimethylꢀ1Hꢀpyrazolꢀ3ꢀyl)ethanone 66 with
the Ellman chiral auxiliary (R)ꢀ2ꢀmethylpropaneꢀ2ꢀsulfinimide.11
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
7

Journal of Medicinal Chemistry
Page 8 of 42
1
2
Scheme 1.a Synthesis of 3ꢀquinoline carboxamides 13ꢀ37 from 6ꢀbromoꢀ4ꢀchloroquinolineꢀ3ꢀ
carboxamide 11.
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) 3ꢀBPinꢀpyridine, Cs2CO3, Pd(PPh3)4, dioxane: water, 100 °C, 5
h (72%); (b) primary amine or cyclopentanethiol (15), DIPEA, DMF (13, 17ꢀ19, 23), NMP (14ꢀ
16, 20-21) or DMA (22ꢀ23), 80ꢀ100 °C; 10ꢀ16 h (19ꢀ85%); (c) 1ꢀmethylꢀ4ꢀBPinꢀ1Hꢀpyrazole,
Cs2CO3, Pd(PPh3)4, dioxane:water, 100 °C, 2 h (96%); (d) 2ꢀethylbutanꢀ1ꢀamine, DIPEA, DMF,
80 °C, 18 h (61%); (e) aryl boronic acid, Cs2CO3, Pd(PPh3)4, dioxane:water, 150 °C, 30 min
(µwave) or 80 °C, 24 h (23ꢀ43%); (f) B2Pin2, (dppf)PdCl2, KOAc, THF, reflux, 3 d (50%); (g)
aryl chloride/bromide, Na2CO3, (dppf)PdCl2, dioxane, 90 °C, 1 h (µwave, 30ꢀ60%); (h) step g,
ACS Paragon Plus Environment
8

Page 9 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
then NaHCO3, Pd/C, MeOH, 25 °C, 16 h (33, 51%); (i) i) PBr3, DCM, reflux, 16 h; ii) NaOMe,
MeOH, 50 °C, 1 h (34%); (j) sodium methanesulfinate, DMA, 40 °C, 30 min (55%).
6
7
8
Scheme 2.a Synthesis of 3ꢀquinoline carboxamides 40 and 43 from ethyl 6ꢀbromoꢀ4ꢀ
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
chloroquinolineꢀ3ꢀcarboxylates.
a
Reagents and conditions: (a) ethylbutyl alcohol, NaH, THF, RT, 2 h (quantitative); (b) i) 3ꢀ
pyridinylboronic acid, Na2CO3, Pd(PPh3)4, dioxane:water, 90 °C, 14 h (54%); ii) NH4OH, H2O,
RT, 14 h (34%); (c) i) NaOH, H2O, 90ꢀ100 °C, 2 h; ii) SOCl2, DMF, 75ꢀ80 °C, 1 h; iii) NH4OH,
H2O/acetone, 90 °C, 10ꢀ15 min (57ꢀ81%); (d) 3ꢀpyridinylboronic acid, Cs2CO3, Pd(PPh3)4,
dioxane:water, 100 °C, 3 h (32%); (e) 2ꢀethylbutanꢀ1ꢀamine, DIPEA, DMA, 100 °C, 2 h (72%).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 42
1
2
Scheme 3.a Synthesis of 5ꢀ, 7ꢀ or 8ꢀ substituted 3ꢀquinoline carboxamides through construction
3
4
5
6
of the quinoline core.
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) ethylꢀ2ꢀcyanoꢀ3ꢀethoxybutꢀ2ꢀenoate (46) or ethyl 2ꢀcyanoꢀ3ꢀ
ethoxyacrylate (47-48), EtOH, 80 °C, 5 h (46, 61%; 47-48, 72%); (b) i) POCl3, 110 °C, 8 h; ii) 2ꢀ
ethylbutanꢀ1ꢀamine or tetrahydroꢀ2Hꢀpyranꢀ4ꢀamine, K2CO3, DMF, RT; 14 h (25ꢀ43%); (c) 3ꢀ
pyridinylboronic acid, Na2CO3, Pd(PPh3)4, dioxane:water, 90 °C, 3 h (42ꢀ66%); (d)
acetaldoxime, Pd(OAc)2, EtOH, 80 °C,
6
h
(9ꢀ14%); (e) aniline, diethyl
(ethoxymethylene)malonate, EtOH, 80 °C, 5 h, then Ph2O, 240 °C, 3 h; (f) i) LiOH,
THF:EtOH:H2O, 55 °C, 1 h; ii) SOCl2, DMF, 70 °C, 3 h; iii) NH4OH, H2O, RT, 10ꢀ15 min (41ꢀ
51% over 4 steps); (g) tetrahydroꢀ2Hꢀpyranꢀ4ꢀamine, DIPEA, DMF, 100 °C; 14 h (25ꢀ86%).
ACS Paragon Plus Environment
10

Page 11 of 42
Journal of Medicinal Chemistry
1
2
Scheme 4.a Final combinations
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
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 42
1
2
a
3
4
Reagents and conditions: (a) iodomethane, NaH, THF, −20 °C, 1 h (92%); (b) B2Pin2,
(dppf)PdCl2, KOAc, THF, reflux, 16 h (quant.); (c) NaOMe, MeOH, −30 °C, 1 h (74%); (d) (1S)ꢀ
1ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀyl)ethanamine, Selectfluor, MeCN, 20ꢀ80 °C, 60 h (25%); (e) (R)ꢀ2ꢀ
methylpropaneꢀ2ꢀsulfinimide, TiOPr4, 2ꢀmethyltetrahydrofuran, 70 °C, 18 h, then sꢀBuLi, −78
°C, 1 h (89%); (f) HCl, dioxane, RT, 1 h (quant.); (g) (1S)ꢀ1ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ3ꢀ
yl)ethanamine or 67 or 68, DIPEA, DMA, 100 °C, 10ꢀ18 h (13ꢀ100%); (h) 63 or 64, Cs2CO3,
Pd(PPh3)4, dioxane:water, 80 °C, 2.5 h (51ꢀ72%).
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
ACS Paragon Plus Environment
12

Page 13 of 42
Journal of Medicinal Chemistry
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
8
A directed subset screen focused on likely kinase inhibitors from the AstraZeneca compound
collection was conducted in a HT29 cellꢀbased radiationꢀactivated ATM assay and trisubstituted
quinolines 4-6 caught our attention amongst an interesting cluster, with either a 2ꢀ
ethylbutylamino or benzylamino side chains at C4 and a primary carboxamide at C3 (Table 1).
Selectivity was assessed using enzyme and cell assays previously reported by our group against a
selection of PI3K and PIKK enzymes.12ꢀ17 Quinoline 4 was found to be a subꢀmicromolar
inhibitor of ATM phosphorylation in cells, and its selectivity profile against members of the
PIKK family and PI3K isoforms was encouraging, with >10ꢀfold enzyme selectivity against
DNAPK and PI3Kα, and >400ꢀfold against mTOR and PI3Kβ, although selectivity against ATR
in cells needed improving (<10ꢀfold). General kinome selectivity for 4 was acceptable for a
screening hit, with only 8 kinases out of 124 tested being inhibited by more than 50% at 1 µM.
Amongst these, we considered GSK3β and KDR to be the most undesirable and concentration
responses were measured (IC50 = 0.55 µM and 8.1 µM respectively). Physicoꢀchemical
properties of 4 were not ideal, notably low solubility (19 µM), moderate to high turnover in
hepatocytes and hERG activity (IC50=2.3 µM), probably the consequence of a rather high
lipophilicity (logD7.4=3.5) – but were thought to be optimizable. When benchmarked against the
known ATM inhibitors 1ꢀ3, our hit 4 looked promising as it combined potency in the cell within
6ꢀfold of 2 and a good selectivity profile. The closely related (1R)ꢀ1ꢀ(4ꢀchlorophenyl)ethanamine
analogue 5 was found to be significantly less potent against ATM but with decreased selectivity
(more potent against ATR, and equipotent against PI3Kα). Interestingly, the 6ꢀ(3’ꢀpyridyl)
analogue 6 proved to gain over 100ꢀfold in ATM cell potency, whilst exhibiting some selectivity
(>10ꢀfold) against ATR, suggesting that the 6ꢀcyano group present in both 4 and 5 was subꢀ
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
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 42
1
2
3
4
5
6
7
8
9
optimal. However, 6 was found to be roughly equipotent against DNAPK and PI3Kα, and
generally even more promiscuous than both 4 and 5, suggesting that the combination of two
pendant aryl groups was undesirable for selectivity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1.a Profile of ATM hit 4-6 vs reported ATM inhibitors 1ꢀ3.
Entry
ATM enzyme IC50
DNAPK enzyme IC50
mTOR enzyme IC50
PI3Kα enzyme IC50
PI3Kβ enzyme IC50
1
2
3
4
5
6
b
<0.002
0.183
5.58
0.458
0.133
>1
0.0008
0.204
6.55
0.452
2.14
>1
0.149
8.50
3.6
0.021
0.879
4.59
1.49
5.44
>1
3.42
>30
2.6
0.008
0.098
3.45
0.247
5.74
>1
0.820
4.40
3.5
0.049
0.109
0.846
0.036
0.139
0.204
14.4
4.70
3.5
0.002
0.005
0.181
0.008
0.021
0.045
0.109
1.90
c
c
c
c
c
PI3Kγ enzyme IC50
b
ATM cell IC50d
1.22
>30
4.0
ATR cell IC50
e
logD7.4
>4.3
<2.7
LLEf
1.9
3.2
2.9
2.6
1.3
Hu/Rat heps Clint
33/230
13/22
4.5/89
12/74
10/41
11/70
(µL/min/106 cells)g
a
All IC50 data are expressed in micromolar (µM) and are the means of at least n=3 (ATM
b
potency) or n=2 (other) independent measurements. ATM enzyme and cell assay protocols are
available in Supplementary Material. Biochemical assays as reported previously.12ꢀ14 d Cellꢀ
c
based assay as reported previously.14
Measured using shakeꢀflask methodology with a
e
f
g
buffer/octanol volume ratio of 100:1. LLE= ATM cell pIC50 − log D7.4. Hepatocyte intrinsic
clearance.18
There are no crystal structures of ATM reported to date, and at the time of this investigation a
homology model of ATM based on PI3Kγ (presenting approximately 19% identity and 36%
similarity in the kinase domain) was used to help rationalize SAR, although most of our
ACS Paragon Plus Environment
14

Page 15 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
molecular designs were driven by experimental data in our ATM cell assay. Fortuitously, a
closely related molecule from the same series (7 – inactive against ATM up to 30 µM in our cell
assay) had previously been crystallized in PI3Kγ, and gave us reasons to believe that 4 could be
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
optimized (Figure 2). We inferred from the crystal structure of 7 in PI3Kγ that the quinoline
nitrogen was the hinge binder, whilst the 6ꢀcyano may be pointing towards the inside of the
pocket as opposed to exposed to solvent. This postulated binding mode was also supported by
unpublished data and later challenged by core substitution (see Investigation of the core section).
The C4 benzylamino side chain of 7 also appeared to be conformationally locked by the 3ꢀ
carboxamide, but conscious of potential alternative binding modes in ATM, we submitted our
hypothesis to the test by testing close analogues of 4-6 from our compound collection.
NH
N
O
N
NH2
7
Figure 2. Crystallised structure of 7 in PI3Kγ (PDB code 5G55), a close sequence homologue of
ATM.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 42
1
2
3
4
Follow-up screening
5
6
7
8
Amongst the most noteworthy C3 analogues found in our collection, quinolines 8 (deletion of
the carboxamide), 9 (carboxylic acid) and 10 (monomethylated carboxamide) were all inactive
against both ATM and ATR in the cell, suggesting a key role for the primary carboxamide in
both providing the required conformation of the 4ꢀsubstitutent and/or in binding to the kinase.
The cell inactivity of the secondary amide 10, despite the removal of one hydrogen bond donor
(HBD) and the increase in lipophilicity, could not be rationalized using our homology model
since it appeared to be positioned close to solvent. This suggested subtle interactions between the
inhibitor and protein which could not be modelled. The apparent optimal nature of the primary
carboxamide in the 3ꢀposition meant that this group was subsequently fixed in future SAR
investigations. It should be noted that despite the presence of two HBDs in this moiety,
carboxamide containing compounds were found to show acceptable permeability and oral
absorption and thus did not hinder further optimization.
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
Table 2. 3ꢀCarboxamide replacements.
Entry
R2
ATM cell ATR cell
a
a
IC50
IC50
H
>30
>30
8
9
>30
>30
>30
>30
10
a All three compounds were tested inactive up to 30 µM in three independent measurements.
ACS Paragon Plus Environment
16

Page 17 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Further followꢀup testing around 4 led us to make the following observations: a) favored C4
analogues were all bearing an NH linker (in line with our internal Hꢀbond hypothesis) and an
extra methylene linker was preferred; b) a potency gain was observed with αꢀsubstituted groups
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
(e.g. benzylamino vs. αꢀmethyl benzylamino), although the preferred stereochemistry was
unclear, as shown in Figure 2; c) evidence showed that 6ꢀaryl groups brought a lot of potency to
the molecules, in line with the 5/6 matched pair.
Figure 2. Followꢀup testing around hit 4. Every marker represents a preꢀexisting collection
molecule tested against a cellꢀbased ATM assay, and molecules are grouped on the x axis by
type of group linker at the C4 position. Matched molecular pairs are connected.
R4
O
R6
NH2
N
With these results in hand, we synthesized and evaluated 3ꢀpyridyl quinoline 13, which
showed a 10ꢀfold improvement in cell potency over our initial hit 4, combined with increased
selectivity over ATR in cells (>150ꢀfold vs <10ꢀfold, Table 3). These results were comparable to
those observed with the 5/6 matched pair, although to a lesser extent, but constituted an
encouraging start. Increased turnover in rat hepatocytes and increased potency against hERG,
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 42
1
2
3
4
5
6
7
8
9
were observed with 13, perhaps expectedly with a logD7.4 of 4.0. The 4ꢀOꢀlinked analogue 40
was made in order to question the intramolecular Hꢀbonding with the 3ꢀcarboxamide which
could still occur but would reverse the orientation of the amide, whilst removing one presumably
masked HBD. The resulting compound proved to be inactive up to 30 µM, again highlighting the
key relationship between the 3ꢀcarboxamide and the 4ꢀamino group. Moreover, 40 had worse
turnover in rat hepatocytes consistent with increased lipophilicity. Similarly, thioether 15 was
made in an attempt to use the oxophilic properties of sulphur in order to conserve the amide
conformation, an effect recently surveyed by Meanwell.19 Compared to matched pair Nꢀlinked
molecule 14, the thioether exhibited c.a. 70ꢀfold loss in cell potency, despite removal of one
HBD and thus disproving our initial hypothesis. Interestingly, 15 proved less lipophilic
(∆logD7.4=−0.4 relative to 14), but displayed increased metabolic instability in rat hepatocytes
suggesting that the sulfur atom may be a metabolic soft spot. In an effort to lower in vitro
intrinsic clearances in both human and rat hepatocytes, replacements for the 2ꢀethylbutylamino
side chain with less lipophilic or less metabolically unstable groups were sought.
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
Investigation at C4
A large exploration at C4 was carried out under the influence of our early observations,
focusing on alkyl groups rather than benzylꢀlike ones to ensure good selectivity of the end
products, and a brief summary is shown in Table 3. 4ꢀCyclopentylamino analogue 14 and 4ꢀ
cyclohexylamino quinoline 16 showed a similar profile to 13, both in terms of potency and
properties. The introduction of an oxygen atom to tetrahydropyranꢀ4ꢀylamino analogue 17
greatly improved the in vitro metabolic profile through a large reduction in lipophilicity
(∆logD7.4=−1.9 relative to 16) but unfortunately the hERG liability remained. 17 was also less
potent than 13 (~13ꢀfold), but LLE was improved by one unit. Cyclohexylmethylamino
ACS Paragon Plus Environment
18

Page 19 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
quinoline 18, which could be seen as the cyclized version of the 2ꢀethylbutylamino parent 13
exhibited a small loss of potency (~6ꢀfold) for no additional benefit. However, tetrahydropyranꢀ
4ꢀylmethylamino analogue 19, despite another drop in potency, showed a significant
improvement in the hepatocyte turnover profile which was similar to that observed with directly
linked tetrahydropyranꢀ4ꢀylamino 17. Introduction of an αꢀmethyl substituent to give
enantiomers 20-21 highlighted a preference for the (S)ꢀenantiomer and brought potency back to
within threeꢀfold of reference 13, whilst retaining acceptable intrinsic clearances in both human
and rat hepatocytes. The same enantiopreference was observed with pyrazole analogues 22 and
23, and furthermore the eutomer 23, despite being less selective against ATR, presented the
advantage of reduced hERG activity (IC50=11 µM) not only compared to its enantiomer but also
closely related analogues. Interestingly, in contrast to other benzylꢀtype compounds such as 6,
compound 23 showed good general kinase selectivity (with the exception of ATR), suggesting
that the nature of the aromatic ring in this position can play a key role in determining the
selectivity profile.
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
19

Journal of Medicinal Chemistry
Page 20 of 42
1
2
Table 3.a Selection of C4 substituents.
3
4
5
6
N
R4
N
O
7
8
9
NH2
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
Entry
R4
Chira ATM cell ATR cell
Hu/Rat
heps CLint
hERG logD7.4 LLE
IC50
b
b
c
d
-lity
IC50
IC50
N/A
0.073
>11
43/140
0.67
4.0
3.1
13
N/A
N/A
N/A
N/A
N/A
>30
0.137
9.7
>30
1.2
42/260
23/65
2.1
0.57
5.1
>4.3
3.6
3.2
3.8
1.9
ꢀ
3.3
1.9
3.1
4.1
40
14
15
16
17
>23
3.2
18/120
34/70
0.132
0.954
3.1
<4.2/8.5
2.1
N/A
N/A
0.456
3.23
3.0
11
40/130
12/19
0.83
>4.3 <2.1
18
19
2.3
2.1
3.4
(R)
(S)
(R)
(S)
0.730
3.5
14/46
14/38
3.4/120
10/31
4.3
2.3
3.8
20
21
22
23
0.168
1.13
>14
>18
0.64
3.6
2.8
11
2.5
2.3
2.5
4.3
3.6
4.4
0.124
a Experimental assays as per Table 1 unless otherwise specified. b All IC50 data are expressed in
micromolar (µM) and are the means of at least n=3 independent measurements. Each has a SEM
± 0.2 log units. Hepatocyte intrinsic clearance expressed in µL/min/106 cells. Inhibition of
c
d
hERG channel in a electrophysiology (IonWorks™) assay.20
ACS Paragon Plus Environment
20

Page 21 of 42
Journal of Medicinal Chemistry
1
2
3
4
Investigation at C6
5
6
7
8
A large investigation at C6 was carried out in parallel to that at C4 with the 2ꢀethylbutylamino
side chain in place at C4, and allowed us to identify interesting aryl groups that are summarized
in Table 4. There seemed to be some degree of tolerance for a range of 6ꢀaryl groups, although
close analogues of the 3ꢀpyridyl such as the 4ꢀpyridyl quinoline 27 or diazines 31-33 were all
less active than parent 13 and displayed equally poor intrinsic clearances. The 3ꢀpyridyl / 4ꢀ
pyridyl matched pair seemed to indicate a key role for the pyridine nitrogen when in the 3ꢀ
position, whilst any additional nitrogen in the ring was disfavored. Interestingly, the ATM
potency of 6ꢀ(1ꢀmethylpyrazolꢀ4ꢀyl)quinoline 25 was within 10ꢀfold of 13, and showed some
improvements of hepatocyte turnover. The real breakthrough however was observed with pꢀ
substituted analogues: 6ꢀ(6ꢀmethylꢀ3ꢀpyridyl)quinoline 34 showed a slight increase in cell
potency compared to the unsubstituted 3ꢀpyridyl, and 6ꢀ[4ꢀ(methoxymethyl)phenyl]quinoline 28
or 6ꢀ[4ꢀ(methylsulfonyl)phenyl]quinoline 29 had remarkably good potencies considering they
lacked the key pyridyl nitrogen. This led us to make the 3ꢀpyridyl analogues 36-37 in the hope of
seeing the combined beneficial effects of the pꢀsubstitution and the 3ꢀpyridyl. This was observed
with quinoline 36 where the introduction of a nitrogen led to an 8ꢀfold cell potency improvement
relative to 28. We were also delighted to observe no activity against ATR in cells up to 30 µM
and much improved kinase selectivity as exemplified by weak or no inhibition of enzymatic
activity of KDR and GSK3β (IC50 = 82 µM and >100 µM respectively). In addition, despite a
high logD7.4, 36 showed a low turnover in human hepatocytes, although it remained high in rat
hepatocytes. Interestingly, the same was not observed with 37, with which only a modest
improvement in potency was achieved. It is also noteworthy that despite a lower logD7.4, the
sulphone showed no improvement in metabolic stability in rat hepatocytes.
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
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 42
1
2
Table 4.a Selection of C6 substituents.
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
Entry
R6
ATM cell ATR cell
Hu/Rat
heps CLint
hERG logD7.4 LLE
IC50
b
b
c
IC50
IC50
0.073
0.525
0.418
>11
7.1
43/140
13/74
37/76
0.67
1.2
4.0
3.4
3.3
3.1
2.7
3.1
13
25
27
0.73
0.62
0.571
0.908
1.840
0.050
>15
2.6
30/93
18/95
2.3
1.6
3.4
3.1
3.3
2.8
2.9
2.4
31
32
33
34
>18
5.5
46/140
7.3/150
0.76
3.3
>4.3 <3.1
0.373
0.070
>30
>30
21/76
18
4.1
3.5
2.3
3.7
28
29
23/210
2.1
0.087
0.045
0.039
>13
>30
>30
150/>300
<6.8/110
15/200
3.1
2.5
1.9
3.7
3.4
35
36
37
>4.3 <3.1
3.1 4.3
a Experimental assays as per Table 3 unless otherwise specified. b All IC50 data are expressed in
micromolar (µM) and are the means of at least n=3 (ATM) or n=2 (ATR) independent
c
measurements. Each has a SEM ± 0.2 log units. Hepatocyte intrinsic clearance expressed in
µL/min/106 cells.
ACS Paragon Plus Environment
22

Page 23 of 42
Journal of Medicinal Chemistry
1
2
3
4
Investigation of the core
5
6
7
8
Further investigation of the core was performed using 13 and 17 as reference compounds.
Simple tetraꢀsubstituted quinolines were synthesized in an attempt to assess the potential for
further substitution on the core and corroborate the postulated binding mode, and the results are
summarized in Table 5. Methyl substitution at C2 exemplified with 2ꢀmethylquinoline 49
resulted in an inactive compound (>30 µM), which could be interpreted either by lack of
necessary space to accommodate the substituent close to the hinge region of the enzyme or
deleterious effect on the conformation of the 3ꢀcarboxamide, in turn affecting that of the 4ꢀ
substitutent. Similarly, 5ꢀfluoro derivative 50 was equally inactive, presumably due to a
detrimental conformational effect on the C4 amine. Interestingly, 50 showed reduced
lipophilicity compared to parent 17 (∆logD7.4=−0.4), which led to decreased activity against
hERG (13 µM). 8ꢀSubstitution by either a fluoro (59) or a chloro (60), also led to inactive
molecules, which was consistent with the postulated binding mode, although the 8ꢀmethyl
quinoline 43 had some residual activity, but was significantly less potent than the parent 13 (6.0
µM, >80ꢀfold). 7ꢀSubstitution proved to be the most interesting, with 7ꢀfluoro analogue 51 being
equipotent to the parent 17, with the advantage of improved selectivity over ATR (>30 µM, >50ꢀ
fold selectivity vs.~ 3ꢀfold selectivity). 7ꢀChloro 58, however, showed a loss of potency (c.a. 3ꢀ
fold) accompanied with an increase in lipophilicity resulting in decreased LLE (2.8 vs 4.1).
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
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 42
1
2
Table 5.a Core substitution.
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
Entry
R4
R2 R5 R7 R8
ATM ATR Hu/Rat hERG logD7.4 LLE
cell
IC50
cell
IC50
heps
CLint
IC50
b
b
c
H
CH3
H
H
H
H
H
H
H
H
H
0.073 >11
>30 >30
6.0 >30
43/140
0.67
0.68
0.36
4.0
2.7
3.1
ꢀ
13
49
43
6/43
CH3
52/89
>4.3 <1.0
H
H
H
H
H
H
H
F
H
H
H
H
H
H
F
Cl
H
H
H
H
H
H
F
0.954
>30 >30
0.606 >30
3.25 >30
>30 >30
>30 >30
3.1 <4.2/8.5
2.1
13
2.0
1.2
10
1.9
1.5
2.1
2.7
2.0
2.1
4.1
ꢀ
4.1
2.8
ꢀ
17
50
51
58
59
60
<3/11
5.5/13
4.3/15
<1/<3
<1/7.3
Cl
1.1
ꢀ
a Experimental assays as per Table 3 unless otherwise specified. b All IC50 data are expressed in
micromolar (µM) and are the means of at least n=3 independent measurements. Each has a SEM
± 0.2 log units. c Hepatocyte intrinsic clearance expressed in µL/min/106 cells.
With the learning of the SAR at each position on the quinoline core, we produced what we
deemed to be the best combinations and some of these are presented in Table 6. Both quinoline
72 and its 7ꢀfluoro analogue 73 proved to be potent cell inhibitors of ATM, with superior
selectivity over ATR (>500ꢀfold) and excellent PIKKꢀfamily selectivity and panꢀkinase
selectivity in both internal (Table 7) and external assays: 72 showed no kinase hit above 30% at
1 µM in a Eurofins panel of 126 kinases and over 1,000ꢀfold enzyme selectivity in a Dundee
panel of 11 lipid kinases (both available in the supplementary information). Figure 3 shows the
evolution of potency and selectivity profiles against a selection of the PI3K and PIKK family
members, GSK3β and KDR from quinoline lead 4.
ACS Paragon Plus Environment
24

Page 25 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Since greater selectivity was achieved with what seemed to be an ATMꢀspecific 6ꢀsubstitutent,
we also revisited benzylꢀlike amino groups at C4 and discovered that pyrazole derivative 74, a
descendant of 23, was similar to 72 in vitro as a potent ATM inhibitor with a remarkably good
selectivity profile: 74 inhibited only 3 kinases out of 386 tested by slightly more than 50% at 1
µM (CLK1, 57%; CLK4, 55%; Mer, 56%). Moreover, 74 presented the advantage over 72 of
inhibiting the hERG ion channel significantly less (22 µM vs 4.5 µM, respectively).
Fluorination of the C6ꢀpyridyl ring (75) was tolerated but showed little impact on either potency
or selectivity. Interestingly, fluorination at this position could be combined with fluorination at
the C7ꢀposition (76) suggesting these two groups do not occupy the same region of space (as we
had initially hypothesized). Further exploration around the pyrazole motif revealed that 4ꢀ
substitution with either fluoride (77) or methyl (78) gave a further slight improvement in cell
potency (ca. 2ꢀ3ꢀfold) for a minimal logD increase and no notable deleterious effect on metabolic
stability or hERG.
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
25

Journal of Medicinal Chemistry
Page 26 of 42
1
2
Table 6.a Best combinations.
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
Entry
R4
R7
R
ATM cell ATR cell
Hu/Rat
heps CLint
hERG logD7.4 LLE
IC50
b
b
c
IC50
IC50
H
F
H
H
0.046
0.052
>30
>30
<2.6/<5.3
<2.7/5.4
4.5
8.9
2.5
2.8
4.8
4.5
72
73
F
H
F
H
F
F
0.033
0.063
0.073
>19
>25
>30
<1.5/5.3
5.7/6.4
4.4/2.3
22
14
24
2.7
2.6
3.0
4.8
4.6
4.1
74
75
76
F
H
0.019
>30
5.5/4.8
>27
3.0
4.8
77
F
H
0.010
>30
3.2/6.9
17
3.0
5.0
78
a Experimental assays as per Table 3 unless otherwise specified. b All IC50 data are expressed in
micromolar (µM) and are the means of at least n=3 independent measurements. Each has a SEM
± 0.2 log units. c Hepatocyte intrinsic clearance expressed in µL/min/106 cells.
ACS Paragon Plus Environment
26

Page 27 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Figure 3. Evolution of the selectivity of ATM inhibitors from lead 4. Every marker represents an
IC50 ratio of selected targets over ATM. Markers are connected and colored by target: enzymatic
ratios for DNAPK, mTOR, PI3Kα, PI3Kβ, PI3Kγ, GSK3β and KDR (left); cellular ratios for
ATR, DNAPK, PI3Kα and PI3Kβ (right).
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
Table 7.a Selectivity data for 72 and 74.
Assay type
Enzyme inhibition
Kinase
ATM
72
74
<0.0012 0.0006
0.643 0.370
DNAPK
mTOR
PI3Kα
PI3Kβ
PI3Kγ
GSK3β
KDR
2.92
2.11 0.445
>27
>1
>100
>100
1.48
12.8
7.1
>100
79
0.046 0.033
Cellular inhibition
ATM
>30
>30
>30 >17.5
>30 >30
>19
18.8
ATR
PI3Kα
PI3Kβ
DNAPK
a All IC50 data are expressed in micromolar (µM).
On the basis of their high solubility, high permeability and low turnover in hepatocytes across
four species, pharmacokinetic parameters were measured for quinolines 72 and 74 (Table 8).
Low to moderate in vivo clearances in rat and dog were observed, with moderate to good
bioavailability in both species. 74 had lower clearance (13 mL/min/kg) and higher bioavailability
(71%) than 72 (50 mL/min/kg and 31%) in dog as predicted from observed differences in vitro
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 42
1
2
in hepatic metabolism (CLint = 2 vs 11 µL/min/106 cells, respectively) and improved
permeability (Papp = 14 and 5.2 x10ꢀ6 cm.sꢀ1 respectively). Furthermore, good exposure was seen
with both 72 and 74, achieving free plasma concentration over the ATM cell IC50 for over 10
hours in mice following doses of 50 mg/kg, where AUC’s of 40 and 33 ꢁM.h and Cmax‘s of 6 and
17 ꢁM were measured respectively. The encouraging pharmacokinetic parameters of 72 and 74
meant that both compounds are suitable for use to evaluate the specific inhibition of ATM in
vivo. Efficacy for 72 will be reported in due course, whilst we report herein efficacy following
treatment with 74 as a representative compound of this series.
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
Table 8. Physical, pharmacokinetic and other properties of 72 and 74.
Parameter
72
74
Solubility (µM)a
590
5.2
69
14
Caco2 A–B Papp (10ꢀ6 cm.sꢀ1)b
Hu/Mu/Rat/Dog PPB (%free)c
Hu/Mu/Rat/Dog heps Clint
(µL/min/106 cells)d
29/24/16/49
18/17/12/18
<2.6/18/<5.3/11 <1.5/28/5.3/2.0
Rat/Dog CL (mL/min/kg)e,f
Rat/Dog Vss (L/kg)e,f
18/50
1.1/2.6
46/31
40
13/13
1.0/1.4
29/71
33
Rat/Dog Bioavailability (%)e,f
Mu AUC0ꢀ24 (ꢀM.h @50mpk)g
a Thermodynamic solubility of solid samples in a 0.1 M aqueous phosphate buffer at pH 7.4 after
24 h at 25 °C.21 b Permeability using Caco2 cells.22 c Plasma protein binding (PPB) was assessed
by equilibrium dialysis in the appropriate species plasma at 37 °C.21 d Hepatocyte intrinsic
e
f
clearance. Male Han Wistar rats were dosed at 2 mg/kg i.v. and 5 mg/kg p.o.. Dogs were
dosed at: 1.7 mg/kg i.v. and 4.2 mg/kg p.o. (72); 1 mg/kg i.v. and 2 mg/kg p.o. (74). g Nude mice
were dosed p.o. with 50 mg/kg of a propylene glycol aqueous solution (72) or 10% DMSO: 90%
captisol solution (74).
Efficacy
3ꢀQuinoline carboxamide 74 was dosed in combination with the topoisomerase I inhibitor
irinotecan (irinotecan: 50 mg/kg, i.p. once on day 1; then 74: 50 mg/kg, p.o. twice daily on days
2ꢀ4; no treatment on days 5ꢀ7) in SW620 (colorectal cell line) tumor bearing,
immunocompromised mice. Irinotecan is the standard of care chemotherapy in second line
ACS Paragon Plus Environment
28

Page 29 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
colorectal cancer and acts through generation of DSBs during DNA replication and activates the
ATM kinase signaling cascade. Previous studies conducted within our laboratories, to be
reported in due course, have shown that monotherapy treatment with a selective ATM inhibitor
does not result in significant efficacy in the same model, which is consistent with the biological
rationale that DNA DSBs are required to first activate ATM, and hence no monotherapy arm was
included in this experiment. From examination of Figure 5, a tolerated combination of irinotecan
with 74 showed significant tumor growth reduction in the SW620 xenograft model after 3 weeks
of the regimen. This reduction was statistically significant and greater than the single agent
treatment with irinotecan, providing evidence of the potentiating effect of 74.
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. In vivo efficacy of 74 in combination with irinotecan in SW620 tumor bearing immune
compromised mice. Days are representative of days post tumor selection. Statistical
significance (p values) was calculated by comparing irinotecan and irinotecan+74 treated groups
using a oneꢀtailed t test. “NS” indicates the tumor volume reduction was not significant from
irinotecan alone.
Control
Irinotecan (50mg/kg IP Day 1 Weekly)
Irinotecan (50mg/kg IP Day 1 Weekly) + 74 (50mg/kg PO Twice daily Days 2-4 Weekly)
2
1.6
1.2
0.8
NS
NS
***
0.4
0
Day 7
Day 14
Day 21
ACS Paragon Plus Environment
29