Highly Potent, Selective, and Orally Bioavailable 4-Thiazol-N-(pyridin-2-yl)pyrimidin-2-amine Cyclin-Dependent Kinases 4 and 6 Inhibitors as Anticancer Drug Candidates: Design, Synthesis, and Evaluation

Article abstract of DOI:10.1021/acs.jmedchem.6b01670

Cyclin D dependent kinases (CDK4 and CDK6) regulate entry into S phase of the cell cycle and are validated targets for anticancer drug discovery. Herein we detail the discovery of a novel series of 4-thiazol-N-(pyridin-2-yl)pyrimidin-2-amine derivatives as highly potent and selective inhibitors of CDK4 and CDK6. Medicinal chemistry optimization resulted in 83, an orally bioavailable inhibitor molecule with remarkable selectivity. Repeated oral administration of 83 caused marked inhibition of tumor growth in MV4-11 acute myeloid leukemia mouse xenografts without having a negative effect on body weight and showing any sign of clinical toxicity. The data merit 83 as a clinical development candidate.

Full text of DOI:10.1021/acs.jmedchem.6b01670

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Highly Potent, Selective and Orally Bioavailable 4-Thiazol-N-(pyridin-2-
yl)pyrimidin-2-amine Cyclin-Dependent Kinase 4 and 6 Inhibitors
as Anticancer Drug Candidates: Design, Synthesis and Evaluation
Solomon Tadesse, Mingfeng Yu, Laychiluh Bantie Mekonnen, Frankie Lam, Saiful
Islam, Khamis Tomusange, Muhammed H. Rahaman, Benjamin Noll, Sunita K.
C. Basnet, Theodosia Teo, Hugo Albrecht, Robert Milne, and Shudong Wang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01670 • Publication Date (Web): 03 Feb 2017
Downloaded from http://pubs.acs.org on February 3, 2017
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 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Highly Potent, Selective and Orally Bioavailable 4ꢀ
ThiazolꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine Cyclinꢀ
Dependent Kinase 4 and 6 Inhibitors as Anticancer
Drug Candidates: Design, Synthesis and Evaluation
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
†
†
†
†
†
Solomon Tadesse , Mingfeng Yu , Laychiluh B. Mekonnen , Frankie Lam , Saiful Islam , Khamis
†
†
†
†
†
Tomusange , Muhammed H. Rahaman , Benjamin Noll , Sunita K. C. Basnet , Theodosia Teo ,
†
†
†∗
Hugo Albrecht , Robert Milne and Shudong Wang
†
Center for Drug Discovery and Development, Sansom Institute for Health Research, School of
Pharmacy and Medical Sciences, and Center for Cancer Biology, University of South Australia,
Adelaide, South Australia 5001, Australia.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Cyclin D dependent kinases (CDK4 and CDK6) regulate entry into S phase of the cell cycle, and
are validated targets for antiꢀcancer drug discovery. Herein we detail the discovery of a novel
series of 4ꢀthiazolꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine derivatives as highly potent and selective
inhibitors of CDK4 and CDK6. Medicinal chemistry optimization resulted in 83, an orally
bioavailable inhibitor molecule with remarkable selectivity. Repeated oral administration of 83
caused marked inhibition of tumour growth in MV4ꢀ11 acute myeloid leukemia mouse
xenografts without having a negative effect on body weight and showing any sign of clinical
toxicity. The data merit 83 as a clinical development candidate.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Page 3 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Cell cycle is an ordered sequence of events that leads to the transition from quiescence or
1
cytokinesis to cell proliferation, and through its checkpoints, ensures genome stability. It
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
comprises four sequential phases viz. S phase, when DNA synthesis occurs; M phase, when the
cell divides into two daughter cells; G1 phase after mitosis and before S phase, when cellular
biosynthetic events take place at high rate and cells grow in size; G2 phase occurring between S
and M phases, during which cells prepare for mitosis. Cells decide to enter S phase or remain
quiescent using signaling pathways that link extracellular cues (e.g. growth factors) to G1 phase
2
of the cell cycle. Progression through G1 phase is regulated by retinoblastoma tumor suppressor
proteins (Rbs), which repress the activity of E2 promoter binding factor (E2F) transcription
factors whose functions are required for transition from G1 to S phase. Phosphorylation of Rb by
G1ꢀphase CDKs releases E2F transcription factors, promoting the transcription of genes that
encode proteins necessary for DNA replication (e.g. cyclin A and cyclin E). The primary kinases
that phosphorylate Rb proteins during G1 phase in mammalian cells are CDK4 and CDK6,
3
hereafter referred to as CDK4/6.
Previous studies have shown that the CDK4/6ꢀRbꢀE2F pathway is disrupted in 90% of
4
ꢀ7
cancers. Besides, genetic studies have established that CDK4/6 are dispensable for the mitotic
cell cycle, evidenced by the facts that mice lacking these kinases are viable and that inactivation
8
,9
of the CDK4/6 genes only affects the proliferation of specific cell types . Thus, CDK4/6
represent logical targets for the development of smallꢀmolecules for therapeutic intervention in
1
0
cancers. Palbociclib (Figure 1), the first and only approved CDK4/6 inhibitor for the treatment
11
of metastatic breast cancer, has rejuvenated the field of selective CDK inhibitors. In addition,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CDK4/6 inhibitors are now becoming recognized as important research tools to answer complex
1
2,13
biological questions.
As most inhibitors bind to the highly conserved ATP binding site, kinase selectivity remains
unconquered. For instance, the in vitro kinase binding profile of the advanced clinical CDK4/6
inhibitors i.e. palbociclib, ribociclib and abemaciclib (Figure 1) against a near kinomeꢀwide
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
4,15
panel disclosed several additional targets.
The offꢀtargets can impair drug safety and
contribute to differential cellular potencies and dissimilar clinical responses. For example,
palbociclib and ribociclib have principally bone marrow toxicities with little toxic effect on
gastrointestinal tract. In contrast, abemaciclib causes abundant gastrointestinal adverse effects.
Palbociclib and ribociclib are dosed intermittently (three weeks on and one week off) while
1
4,15
abemaciclib is administered continuously due to a relatively low incidence of neutropenia.
Jointly, these lines of evidence suggest that new pharmacophores might afford the desired kinase
selectivity profiles and thus offer an effective way to minimize undesirable side effects.
Figure 1. Structures of the advanced clinical CDK4/6 inhibitors.
16ꢀ18
In continuation of our search for new and selective CDKꢀtargeting drugs,
herein we present
the discovery of potent, highly selective, and orally bioavailable CDK4/6 inhibitor 83. We
demonstrate 83 exhibits significant antiꢀcancer efficacy in an MV4ꢀ11 acute myeloid leukemia
(AML) subcutaneous xenograft mouse model.
4
ACS Paragon Plus Environment

Page 5 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS AND DISCUSSION
Rational Drug Design. Due to evolutionary conservation of the ATPꢀbinding site, the
discovery of highly selective smallꢀmolecule ATPꢀantagonistic CDK inhibitors is a daunting
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
10,19ꢀ21
task.
In a bid to identify both potent and selective ATP competitive CDK2 or CDK9
inhibitors, we have previously reported the discovery of the 2ꢀanilinoꢀ4ꢀ(thiazolꢀ5ꢀyl)pyrimidine
1
6ꢀ18
pharmacophore (Figure 2).
Figure 2. Structural modifications to 2ꢀanilinoꢀ4ꢀ(thiazolꢀ5ꢀyl)pyrimidine pharmacophore.
Our previous studies on exploration of structureꢀactivity relationships (SARs) of 2ꢀanilinoꢀ4ꢀ
(
thiazolꢀ5ꢀyl)pyrimidines focused on modifications to the aniline and thiazole moieties of the
pharmacophore. A large number of 2ꢀanilinoꢀ4ꢀ(thiazolꢀ5ꢀyl)pyrimidine analogues were
synthesized and screened against a wide panel of CDKs, and the results thus obtained disclosed
1
6ꢀ18
several compounds with modest CDK4 inhibitory potency.
incorporation of a geometrically fitting, positively ionizable group into the 2ꢀanilinoꢀ4ꢀ(thiazolꢀ
ꢀyl)pyrimidine scaffold would result in favorable interaction(s) with the acidic residue(s) Asp99
We also proposed that
5
2
2
and/or Glu144 in the ATP binding site of CDK4 (Figure 3). As hypothesized, amineꢀ
substituted anilines positioned at C2 of the pyrimidine ring afforded moderately potent and
relatively selective CDK4 inhibitors. In addition, methylation of the bridging NH between the
phenyl and pyrimidine moieties led to an inactive compound. This was not surprising because
this bridging NH and pyrimidinylꢀN1 act as a key hydrogen bond donor and acceptor to interact
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with the backbone carbonyl and the NH groups of Val in the kinase hinge region, respectively. In
another relevant study, it was also speculated that the interactions with the relatively less
2
conserved hinge regions of CDK4/6 would be enhanced by the introduction of aromatic sp
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
nitrogen atom near to the rarely conserved His95 in CDK4 (His100 in CDK6). In fact,
Toogood et al. have shown that aminoꢀpyridinyl compounds possess exquisite selectivity
2
4
towards CDK4/6 over CDK2 than do their anilino counterparts. More recently, the pyridinyl
nitrogen atom of abemaciclib has been shown to interact with the imidazole ring of His100 of
CDK6 via a water bridge, and such an interaction is also possible in the cases of palbociclib and
15
ribociclib due to the three molecules sharing the common pyrimidineꢀNHꢀpyridine motif.
Based on this knowledge, we explored whether structural modifications to the 2ꢀanilinoꢀ4ꢀ
(thiazolꢀ5ꢀyl)pyrimidine pharmacophore would lead to the generation of potent CDK4/6ꢀspecific
inhibitors. To this end, we replaced the phenyl moiety with a pyridine ring, incorporated a
variety of ionizable groups, and introduced an electron withdrawing group at C5 position of the
pyrimidine ring. On the other hand, we retained some structural features that were shown to
contribute towards the potency of the pharmacophore. The pyrimidine ring, which by virtue of its
N1 and N3 is expected to make a hydrogen bond with Val of the hinge region of CDK4/6 and to
ensure coꢀplanarity of the pyrimidine and pyridine rings (an essential feature for ATP mimetic
kinase inhibitors), respectively, was retained. Our previous studies showed that the thiazole C2ꢀ
amino site interacted strongly with the Asp145 residue of the DFG motif of CDK2 (Asp167 in
CDK9), enhancing the hydrophobic interactions of the thiazolꢀ4ꢀyl methyl group with the Phe80
1
6ꢀ18,22
gatekeeper residues of CDK2 (Phe103 in CDK9),
and these interactions are expected with
CDK4/6 because the DFG motif and the gatekeeper Phe are highly conserved in the CDK family.
As a result, the thiazole C2ꢀamino moiety was not significantly altered. A total of 37 derivatives
6
ACS Paragon Plus Environment

Page 7 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 4ꢀ(4ꢀmethylthiazolꢀ5ꢀyl)ꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine (Figure 2) were synthesized
accordingly and their CDK inhibitory activities and antiꢀproliferative effects are summarized in
Tables 1ꢀ4.
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. ATP binding site of CDK6 in complex with the inhibitor palbociclib (PDB ID: 2EUF)
aligned with that of CDK4 (PDB ID: 2W96). CDK6 is shown in brown and CDK4 in cyan.
Palbociclib, shown in green and bound in the ATP binding pocket of the kinases, forms two
hydrogen bonds with conserved Val101 and one hydrogen bond with conserved Asp163. The
hydrogen bonds are shown in black dashed lines. His95, Asp99, Thr102 and Glu144 of CDK4
are shown in cyan sticks. The figure was prepared using PyMOL1.3 (Schrödinger Inc., 2013).
Synthetic Chemistry. The synthesis of 4ꢀthiazolꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine
derivatives 78ꢀ112, 115 and 116 requires the preparation of two pairs of building blocks, i.e.,
enaminones/guanidines and pyrimidinꢀ2ꢀamines/aryl bromides. The synthetic route to a variety
of enaminones and two pyrimidinꢀ2ꢀamines is described in Scheme 1.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Isothiocyanatocyclopentane
2
was synthesized from the commercially available
2
5
cyclopentylamine 1 using a procedure reported by Munch et al. Treatment of isothiocyanates 2
or 3 (obtained from commercial source) with ammonia afforded Nꢀsubstituted thioureas 4ꢀ6, with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
26
N,Nꢀdisubstituted thiourea 5 isolated as a byꢀproduct in the synthesis of 2. We speculate that 5
is formed from dicyclopentylcarbamodithioate, which is formed by the reaction between
cyclopentyl(dithiocarboxy)amide intermediate and a halocyclopentane (minor contaminant of
cyclopentylamine starting material). The dicyclopentylcarbamodithioate can then be
desulfurylated by reacting with Boc O and ammonia to yield 5. 5ꢀAcetylthiazoles 8ꢀ12 were
2
prepared by the reaction of 3ꢀchloroꢀ2,4ꢀpentanedione with the prepared thioureas 4ꢀ6 or the
commercially available isopropyl thiourea 7 using HantzschꢀTraumann thiazole synthesis
2
7
method with the single exception of analogue 9 which was made from a twoꢀstep reaction of
1,1,1ꢀtrifluoropentaneꢀ2,4ꢀdione with [hydroxy(tosyloxy)iodo]benzene and 1ꢀ
2
8
cyclopentylthiourea 4, sequentially. 5ꢀAcetylthiazoles 8ꢀ12 were readily converted to the
corresponding enaminones 13 and 15ꢀ20 either by refluxing with N,Nꢀdimethylformamide
29
dimethyl acetal (DMFꢀDMA) or using tertꢀbutoxybis(dimethylamino)methane (Bredereck’s
1
8,29
reagent) as previously reported.
Enaminone 13 was fluorinated with SelectFluor in methanol
1
7
on an ice bath to give fluoride 14. The synthesis of pyrimidinꢀ2ꢀamines 21 and 22 was achieved
by the coupling of enaminones 13 and 14 with guanidine hydrochloride, a classical route to
3
0
construct pyrimidine ring systems.
8
ACS Paragon Plus Environment

Page 9 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 1. Synthesis of enaminones 13ꢀ20 and pyrimidinꢀ2ꢀamines 21 and 22
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) CS , NEt , Boc O, DMAP, EtOH, rt, 30 min, ice bath, 15 min, rt,
2
3
2
3
0 min; (b) NH , H O, MeOH, rt, 24 h; (c) 3ꢀchloroꢀ2,4ꢀpentadione, pyridine, MeOH, rt, 4ꢀ48 h;
3 2
or 1,1,1ꢀtrifluoropentaneꢀ2,4ꢀdione, hydroxyl(tosyloxyl)iodobenzene, MeCN, reflux, 1 h; (d)
DMFꢀDMA, reflux, 3ꢀ36 h; or microwave, 150 °C, 30 min; or Bredereck’s reagent, reflux, 6 h;
(e) SelectFluor, MeOH, ice bath, 4 h; (f) 13 or 14, guanidine hydrochloride, NaOH, 2ꢀ
b
c
methoxyethanol, reflux, 6 h. Intermediates 3 and 7 were obtained from commercial sources.
most likely from dicyclopentylcarbamodithioate, which is formed by the reaction between
cyclopentyl(dithiocarboxy)amide intermediate and a halocyclopentane (minor contaminant of
cyclopentylamine starting material) .
The synthesis of a wide range of guanidines is depicted in Scheme 2. Nucleophilic substitution
of 5ꢀbromoꢀ2ꢀnitropyridine with various Nꢀheterocycles 23ꢀ32 in dimethylsulfoxide furnished
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
31
nitro compounds 33, 34, 36, 37, and 39ꢀ44 in moderate to excellent yields. The secondary
amino group of 34 and 37 was found to interfere with the subsequent guanidine formation with
N,N’ꢀbisꢀBocꢀSꢀmethylisothiourea, and therefore was masked by the tertꢀbutyloxycarbonyl (Boc)
group to give intermediates 35 and 38, respectively. Nitro compounds 33, 35, 36 and 38ꢀ44 were
reduced with gaseous hydrogen in the presence of 10% Pd/C in methanol to give respective
amines 45ꢀ54 in good to excellent yields (82ꢀ100%), each of the amines 45ꢀ54 and the
commercially available pyridineꢀ2ꢀamine 55 were subsequently coupled with N,N’ꢀbisꢀBocꢀSꢀ
methylisothiourea in the presence of Et N and HgCl to afford bisꢀBocꢀprotected guanidine
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2
32
derivatives 56ꢀ66. Removal of Boc groups from 56ꢀ66 was carried out in a mixture of TFA and
DCM (1:1) at reflux, giving the corresponding guanidines 67ꢀ77 in quantitative yields. The last
building block 1ꢀ((6ꢀbromopyridinꢀ3ꢀyl)methyl)ꢀ4ꢀethylpiperazine 114 was prepared in 25%
yield by reductive amination of 6ꢀbromopyridineꢀ3ꢀcarbaldehyde 113 with ethylpiperazine using
sodium triacetoxyborohydride (Scheme 3).
Finally, all desired 4ꢀthiazolꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine derivatives were obtained
16
through either pyrimidine ring cyclization between an enaminone and a guanidine (78ꢀ112,
3
3
Scheme 2) or Buchwald−Hartwig amination of a bromide and an amine (115 and 116, Scheme
). Each enaminone (13ꢀ20) was condensed with an appropriate pyridinyl guanidine (67-77)
3
17
under our previously reported microwaveꢀassisted conditions, giving 78ꢀ112 in poor to
moderate yields (5ꢀ57%) (Scheme 2). In parallel, 115 and 116 were prepared by the palladiumꢀ
catalyzed coupling of pyrimidinylꢀ2ꢀamine 21 or 22 with bromide 114 (Scheme 3).
10
ACS Paragon Plus Environment

Page 11 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 2. Synthesis of guanidine derivatives 67ꢀ77 and 4ꢀthiazolꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀ
a
amines 78-112
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) 5ꢀbromoꢀ2ꢀnitropyridine, Et N, DMSO, 120 °C, 16 h; (b) Boc O,
3
2
DMAP, Et N, DCM, o/n, rt; (c) H , 10% Pd/C, MeOH, o/n, rt; (d) N,N′ꢀbisꢀBocꢀSꢀ
methylisothiourea, HgCl , Et N, DCM, 0 °C to rt, 14 h; (e) TFA/DCM (1:1), reflux, 16 h; (f) 13ꢀ
0, NaOH, 2ꢀmethoxyethanol, microwave, 180 °C, 1 h. No substitution on carbon 5 of the
pyridine ring and obtained from commercial source. No substitution on carbon 5 of the pyridine
3
2
2
3
b
2
c
d
ring. See tables 1, 2 and 4 for the structures of 78ꢀ112.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 3. Synthesis of 115 and 116 via Buchwald−Hartwig amination
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) ethylpiperazine, sodium triacetoxyborohydride, DCM, 0 °C to rt,
2 h; (b) 21 or 22, Pd (dba) , xantphos, tꢀBuONa, 1,4ꢀdioxane, microwave, 150 °C, 1 h.
1
2 3
12
ACS Paragon Plus Environment

Page 13 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. Structures and in vitro biological activities of 4ꢀ(4ꢀmethylthiazolꢀ5ꢀyl)ꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine derivatives 78-99
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
a
b
Compd Substituent
CDK inhibition, K (µM)
Growth inhibition, GI (µM)
50
i
1
2
5
c
c
c
c
c
c
No.
R
R
R
4D
6D3
1B
2A
7H
9T1
MV4ꢀ11
MDAꢀMBꢀ453
0.070 ± 0.013
0.102 ± 0.117
0.130 ± 0.011
0.287 ± 0.070
0.591 ± 0.256
3.683 ± 0.285
0.055 ± 0.012
7.035 ± 0.710
0.215 ± 0.052
7
8
cꢀPent
cꢀPent
cꢀPent
cꢀPent
cꢀPent
cꢀPent
cꢀPent
cꢀPent
cꢀPent
H
H
H
H
H
H
H
H
H
NH
CH
0.001 0.034 > 5
> 5
1.108 0.220 0.023 ± 0.024
79
2
NH
0.001 0.008 4.070 0.278 0.282 0.508 0.029 ± 0.002
0.002 0.009 2.675 0.206 0.865 0.180 0.009 ± 0.000
0.002 0.011 3.020 0.355 0.780 0.141 0.009 ± 0.001
0.006 0.093 3.252 0.776 3.453 0.286 0.024 ± 0.028
8
0
1
NCH
3
2
8
NCH
CH
3
82
NCOCH
3
8
3
4
O
0.004 0.030 NA
0.003 0.133 1.230 0.181 1.187 0.173 0.013 ± 0.005
0.070 0.257 NA > 5 NA > 5 0.191 ± 0.029
4.683 NA
> 5
0.209 ± 0.030
8
CHNH
CH
NSO
2
85
2
8
6
2
CH
3
0.007 0.055 1.100 0.077 2.640 1.321 0.176 ± 0.009
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
d
8
8
8
9
7
8
9
0
cꢀPent
H
NSP
NH
0.570
ꢀ
ꢀ
ꢀ
ꢀ
> 5
0.300 ± 0.035
4.379 ± 0.691
0.402 ± 0.006
3.360 ± 0.286
6.864 ± 0.798
0.373 ± 0.117
0.279 ± 0.044
0.494 ± 0.081
0.150 ± 0.029
0.031 ± 0.002
0.938 ± 0.068
0.618 ± 0.193
0.344 ± 0.126
0.115 ± 0.024
0.326 ± 0.047
cꢀPent CH
cꢀPent CH
cꢀPent CH
0.002 0.010 1.39
0.002 0.010 > 5
0.010 0.031 3.17
0.071 0.539 > 5
0.005 0.066 2.04
0.019 0.485 NA
0.026 0.100 > 5
0.174 3.20
0.476 > 5
0.121 NA
1.801 0.285 ± 0.041
1.800 0.020 ± 0.015
3
3
3
NCH
NCOCH
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
> 5
> 5
0.328 ± 0.007
0.714 ± 0.179
3
91
cꢀPent cꢀPent NH
> 5
> 5
9
2
3
Ph
H
NH
> 5
1.66
NA
> 5
0.436 0.056 ± 0.011
9
Ph
H
NCH
NCH
NH
3
NA
1.04
> 5
0.508 ± 0.042
0.421 ± 0.044
94
Ph
CH
3
3
2.00
95
96
97
98
99
iꢀPr
iꢀPr
iꢀPr
iꢀPr
iꢀPr
H
0.005 0.011 3.140 0.240 0.775 2.420 0.093 ± 0.010
0.016 0.028 > 5 0.800 1.16 0.925 0.255 ± 0.085
H
CH
NH
2
H
NCH
0.003 0.015 3.815 0.399 0.760 0.773 0.075 ± 0.005
0.021 0.105 2.695 0.200 4.385 3.717 2.071 ± 0.321
3
H
NCOCH
O
3
H
0.041 0.082 > 5
0.003 0.027 > 5
0.127 > 5
> 5 > 5
> 5
0.032 ± 0.003
Palbociclib
0.364 0.050 ± 0.004
a
K values were used as ATP concentrations. Apparent inhibition constants (K ) were calculated using the half maximal inhibition
m
i
(IC ) and the appropriate K (ATP) of each kinase as described in the experimental section. NA indicates no inhibition or compound
5
0
m
1
4
ACS Paragon Plus Environment

Page 15 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
b
activity that could not be fit to an IC curve. Antiꢀproliferative activity was determined by 72 h resazurin and MTT assays using
5
0
c
MV4ꢀ11 and MDAꢀMBꢀ453 cell lines, respectively. The data given are derived from at least two replicates. 4D1, 6D3, 1B, 2A, 7H,
and 9T1 represent CDK4ꢀcyclin D1, CDK6ꢀcyclin D3, CDK1ꢀcyclin B, CDK2ꢀcyclin A, CDK7ꢀcyclin H and CDK9ꢀcyclinT1,
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
d
respectively. NSP indicates no substitution at C5 of the pyridine ring.
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Structure-Activity Relationship Analysis. The very first 4ꢀ(4ꢀmethylthiazolꢀ5ꢀyl)ꢀNꢀ
(
pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀamine prepared with an ionizable piperazine group on the pyridine
ring was 78. This molecule was found to exhibit excellent potency (K = 1 and 34 nM for CDK4
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and CDK6, respectively) and a relatively high selectivity for CDK4/6 over CDK1, CDK2, CDK7
and CDK9. Compound 78 is also a highly effective antiꢀproliferative agent with a GI50
(
concentration for 50% inhibition of cell proliferation) value of 23 nM against MV4ꢀ11. As
expected, inhibition of CDK4/6 by 78 resulted in accumulation of MV4ꢀ11 cells in the G1 phase
of cell cycle in a concentration dependent manner (Figure 4). At the highest concentration (0.40
µM) tested, 78 accumulated 85% G1 population compared to the 61% in untreated cells. Similar
cell cycle effects of palbociclib were also observed. This observation encouraged us to further
explore the utility and versatility of the 4ꢀ(4ꢀmethylthiazolꢀ5ꢀyl)ꢀNꢀ(pyridinꢀ2ꢀyl)pyrimidinꢀ2ꢀ
amine scaffold as CDK4/6 inhibitors.
16
ACS Paragon Plus Environment

Page 17 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 4. Cell cycle analysis of 78. Compound 78 induced G1 cell cycle accumulation in MV4ꢀ
11 cells. The cells were treated with increasing concentrations of 78 or palbociclib (Palb) for 24
h, and analyzed by flow cytometry. The experiments were repeated twice and representative data
were presented.
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
Several structural alterations in relation to the substitution at the amino site of the thiazole
moiety were made to examine whether the nature of the substituent would affect enzymatic and
cellular efficacies (Table 1). Replacement of the cyclopentyl group (78 and 80) with the bulkier
and more rigid aromatic phenyl group (92 and 93) resulted in a moderate reduction in both
CDK4/6 and cellular activity. Similarly, substitution with the smaller and more flexible isopropyl
group (78ꢀ80 and 82 versus 95ꢀ98) slightly decreased both CDK inhibitory activity and antiꢀ
proliferative effect. However, despite its weaker CDK4/6 inhibitory activity (K = 41 and 82 nM
i
for CDK4 and CDK6, respectively), 99 with an isopropyl substituent was more potent than the
corresponding cyclopentyl analog 83 (K = 4 and 30 nM for CDK4 and CDK6, respectively) in
i
terms of antiꢀproliferative activity (GI = 32 and 115 nM for 99 versus GI = 209 nM and > 3
5
0
50
µM for 83 against MV4ꢀ11 and MDAꢀMBꢀ453, respectively), suggesting offꢀtarget effect(s) by
9. Increasing the steric bulk at the amino site by introduction of a second alkyl group, i.e.,
9
methyl or cyclopentyl, moderately to greatly diminished the potency against both cell lines,
which could be at least partially attributed to the reduced CDK7/9 inhibitory activity observed
(78, 80, and 82 versus 88ꢀ90, and 78 versus 91). Collectively, structural modifications made to
the amino site of the thiazole ring indicated that monoꢀsubstitution of a cyclopentyl group
afforded the optimal combination of both enzymatic and cellular potencies as well as the
selectivity for CDK4/6. As a result, this functionality was retained and further alterations at C4
position of the thiazole ring were carried out (Table 2). Substitution of the methyl group with the
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
bulkier electron withdrawing trifluoromethyl (78, 80, 82 and 83 versus 100ꢀ103) markedly
increased the potency against MV4ꢀ11 cells with GI values ranging from 1 to 69 nM while the
50
trend was not conclusive for MDAꢀMDꢀ453 cells. In addition to the increased CDK2/7 inhibitory
activity, the inhibition of CDK1, prevalently considered to induce toxicity against both malignant
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
4
and untransformed cells, was found to be significantly enhanced, thus severely jeopardizing the
CDK4/6 selectivity profile. Taken together, the data show that the cyclopentylamino and methyl
group are optimal at C2 and C4 positions of the thiazole ring, respectively.
Table 2. Structures and biological activities of Nꢀcyclopentylꢀ5ꢀ(2ꢀ(pyridinꢀ2ꢀ
ylamino)pyrimidinꢀ4ꢀyl)ꢀ4ꢀ(trifluoromethyl)thiazolꢀ2ꢀamine derivatives 100-104
a
b
Compd. Substituent
CDK inhibition, K (µM)
Growth inhibition, GI (µM)
i
50
5
c
c
c
c
c
c
No.
R
4D1
6D3
1B
2A
7H
9T1
MV4ꢀ11
MDAꢀMBꢀ453
100
NH
NCH
0.008 0.002 0.220 0.022 0.194 0.258 0.002 ± 0.001 0.081 ± 0.039
0.001 0.003 0.241 0.022 0.189 0.831 0.001 ± 0.001 0.420 ± 0.120
1
1
1
1
01
02
03
04
3
NCOCH
O
0.004 0.006 0.297 0.014 2.615 > 5
0.008 0.011 NA 0.154 NA > 5
0.011 0.007 0.345 0.015 1.900 > 5
0.012 ± 0.001 0.077 ± 0.001
0.069 ± 0.013 0.415 ± 0.103
0.004 ± 0.001 0.336 ± 0.188
3
NCHO
a
Km values were used as ATP concentrations. Apparent inhibition constants (K ) were
i
calculated using the half maximal inhibition (IC ) and the appropriate K (ATP) of each kinase
5
0
m
as described in the experimental section. NA indicates no inhibition or compound activity that
18
ACS Paragon Plus Environment

Page 19 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
b
could not be fit to an IC curve. Antiꢀproliferative activity was determined by 72 h resazurin
5
0
and MTT assays using MV4ꢀ11 and MDAꢀMBꢀ453 cell lines, respectively. The data given are
c
derived from at least two replicates. 4D1, 6D3, 1B, 2A, 7H, and 9T1 represent CDK4ꢀcyclin D1,
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
CDK6ꢀcyclin D3, CDK1ꢀcyclin B, CDK2ꢀcyclin A, CDK7ꢀcyclin H and CDK9ꢀcyclinT1,
respectively.
With the optimal thiazole moiety in hand, we next explored a variety of substituents at C5
position of the pyridine ring. Given that a basic amino substituent at this position was previously
2
4
shown to improve both kinase selectivity and pharmaceutical properties, our effort mainly
focused on introduction of various nitrogenꢀcontaining cyclic substituents. In fact, analysis of the
crystal structure of palbociclib bound to CDK6 revealed that the piperazine ring sits in the kinase
channel near the solvent accessible region, and thus is capable of accommodating a wide variety
3
5
of substituents (Figure 3). The secondary amino site of the piperazine ring is expected to be
protonated at physiological pH and consequently interact with Thr102 in CDK4 (Thr107 of
CDK6) which corresponds to the unfavorable (due to electrostatic repulsion) positively charged
Lys residues of CDK1/2, rendering the selectivity towards CDK4/6 over these kinases. In
contrast, in CDK9, the corresponding residue is Gly which can be in contact with the protonated
amino site, and thus contribute to CDK9 inhibition, albeit with a higher inhibition constant. In
our current study, extending the cyclic secondary amino site of 78 further into the solvent
channel gave rise to the exocyclic primary amino compound 84, which exhibited decreased
potency for CDK6 and selectivity for CDK4 over CDK2, and a marginally increased activity
against both cell lines (Table 1). The piperidine analogue 85, lacking the secondary amino site,
was 70ꢀ and > 7ꢀfold less potent against CDK4 and CDK6, respectively, than 78 (K = 70 and
i
2
57 nM for 85 versus K = 1 and 34 nM for 78 towards CDK4 and CDK6, respectively). The
i
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
absence of a heterocyclic substituent on the pyridine ring (87) was not tolerated with a K value
i
of 570 nM for CDK4. The significant loss of the CDK4 inhibitory potency in the cases of 85 and
87 confirmed the requirement of an appropriate heterocyclic substituent at the C5 position of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
pyridine ring for CDK4/6 inhibition. Nevertheless, these two compounds exhibited good antiꢀ
proliferative activity against MV4ꢀ11 cells, indicating potential offꢀtargets. Elongation of the
linker by insertion of a methylene spacer between the pyridine ring and the piperazine moiety
(115 and 116 versus 81 and 108) did not significantly impact the CDK4/6 selectivity and cellular
potency (Tables 1, 3 and 4).
Table 3. Structures and in vitro biological activities of Nꢀcyclopentylꢀ5ꢀ(2ꢀ((5ꢀ((4ꢀ
ethylpiperazinꢀ1ꢀyl)methyl)pyridinꢀ2ꢀyl)amino)pyrimidinꢀ4ꢀyl)ꢀ4ꢀmethylthiazolꢀ2ꢀamine
derivatives 115 and 116
a
b
Compd. Substituent
CDK inhibition, K (µM)
Growth inhibition, GI (µM)
i
50
4
c
c
c
c
c
c
No.
15
116
R
H
F
4D1
6D3
1B 2A
7H
9T1 MV4ꢀ11
MDAꢀMBꢀ453
1
0.006 0.009 > 5 0.416 0.211 1.98 0.019 ± 0.003 0.069 ± 0.023
0.003 0.014 NA 0.620 0.630 3.57 0.010 ± 0.002 0.622 ± 0.208
a
Km values were used as ATP concentrations. Apparent inhibition constants (K ) were
i
calculated using the half maximal inhibition (IC ) and the appropriate K (ATP) value of each
5
0
m
kinase as described in the experimental section. NA indicates no inhibition or compound activity
b
that could not be fit to an IC50 curve. Antiꢀproliferative activity was determined by 72 h
resazurin and MTT assays using MV4ꢀ11 and MDAꢀMBꢀ453 cell lines, respectively. The data
20
ACS Paragon Plus Environment

Page 21 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
c
given are derived from at least two replicates. 4D1, 6D3, 1B, 2A, 7H, and 9T1 represent CDK4ꢀ
cyclin D1, CDK6ꢀcyclin D3, CDK1ꢀcyclin B, CDK2ꢀcyclin A, CDK7ꢀcyclin H and CDK9ꢀ
cyclinT1, respectively.
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
Recently, we successfully optimized the pharmacokinetic (PK) properties of CDK9 inhibitors
3
6,37
by introduction of a fluorine atom at C5 position of the pyrimidine ring.
This strategy was
also deployed in the current study to seek potent and selective CDK4/6 inhibitors with improved
bioavailability. In addition, we envisioned that the electronegative fluorine atom would acidify
the bridging NH between the pyrimidine and pyridine rings, thereby increasing the strength of
3
8
the hydrogen bonding with the carbonyl group of Val in the hinge region. Regrettably, all the
attempts thus made led to little success in further improving CDK4/6 inhibitory activity and
selectivity, antiꢀproliferative potency (Table 4), and PK properties (data not shown).
Nonetheless, probably due to its bulkier ionizable amino site in the vicinity of Thr102, the 4ꢀ
(
dimethylamino)piperidine analogue 112 possesses an increased selectivity towards CDK4.
Interestingly, all morpholineꢀcontaining compounds 83, 99, 103 and 110 displayed excellent
selectivity towards CDK4/6 over the other CDKs tested (Tables 1, 2 and 4). Particularly, 83
showed > 1,200ꢀfold selectivity over CDK2 and was inactive against the rest of CDKs tested.
Setting a maximum K value of 10 nM and a minimum selectivity of 100ꢀfold for CDK4 as
i
criteria, we have identified 9 compounds, i.e., 78, 79, 83, 89, 97, 107, 108, 112 and 116, as
highly selective CDK4/6 inhibitors, all of which exhibited potent cellular antiꢀproliferative
activities in cancer cell lines.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 4. Structures and in vitro biological activities of Nꢀcyclopentylꢀ5ꢀ(5ꢀfluoroꢀ2ꢀ(pyridinꢀ2ꢀ
ylamino)pyrimidinꢀ4ꢀyl)ꢀ4ꢀmethylthiazolꢀ2ꢀamine derivatives 105-112
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
b
Compd. Substituent
CDK inhibition, K (µM)
Growth inhibition, GI (µM)
i
50
5
c
c
c
c
c
c
No.
R
4D1
6D3
1B
2A
7H
9T1
MV4ꢀ11
MDAꢀMBꢀ453
1
1
05
06
NH
0.003 0.007 2.235 0.256 0.790 0.787 0.053 ± 0.004 0.780 ± 0.598
0.014 0.010 > 5 0.417 0.815 0.679 0.029 ± 0.002 0.083 ± 0.009
0.001 0.003 2.330 0.103 2.020 0.505 0.073 ± 0.028 0.202 ± 0.030
CH NH
2
107
NCH
3
2
108
119
110
111
NCH
CH
3
0.002 0.006 NA
0.034 0.023 > 5
0.349 0.685 > 5
0.013 ± 0.002 0.066 ± 0.019
NCOCH
O
0.228 > 5 4.990 0.015 ± 0.002 8.107 ± 1.147
3
0.006 0.020 4.350 0.104 > 5
> 5
> 5
0.290 ± 0.062 1.437 ± 0.304
2.628 ± 0.582 2.813 ± 0.089
NSO
CHN(CH
CH
0.039 0.101 NA
0.001 0.031 NA
0.348 NA
1.150 NA
2
3
112
3
)
2
1.091 0.454 ± 0.040 0.356 ± 0.024
a
Km values were used as ATP concentrations. Apparent inhibition constants (K ) were
i
calculated using the half maximal inhibition (IC ) and the appropriate K (ATP) of each kinase
50
m
as described in the experimental section. NA indicates no inhibition or compound activity that
b
could not be fit to an IC curve. Antiꢀproliferative activity was determined by 72 h resazurin
50
and MTT assays using MV4ꢀ11 and MDAꢀMBꢀ453 cell lines, respectively. The data given are
c
derived from at least two replicates. 4D1, 6D3, 1B, 2A, 7H, and 9T1 represent CDK4ꢀcyclin D1,
CDK6ꢀcyclin D3, CDK1ꢀcyclin B, CDK2ꢀcyclin A, CDK7ꢀcyclin H and CDK9ꢀcyclinT1,
respectively.
ACS Paragon Plus Environment
22

Page 23 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Kinase Selectivity Profiling of 83. Upon identification of 9 potent compounds with high
selectivity for CDK4/6 over the other CDK family members tested, a selection of compounds,
i.e., 78, 83, 107 and 112, were further screened against a panel of 50 different kinases (Table S1).
Compounds 78 and 107 potently inhibited (< 10% kinase residual activity at 10 µM) a number of
kinases including CK1δ, DAPK1, DYRK1A/B, FGFR1, Flt1, Flt3, Flt4, HIPK2, JNK3, Mnk2,
Pimꢀ1 and SGK. In contrast, DYRK1A/B and HIPK2 were the only two other kinases
significantly affected by 112. Strikingly, 83 was the most selective CDK4/6 inhibitor with no
appreciable effect on other kinases tested. The subsequent profiling of 83 against a wider panel
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
of 369 kinases at the concentration of 1 µM (~ 250 × K ) (Figure 5, Table S2) revealed that 83 is
i
exceptionally selective for CDK4/6 within the human kinome; DYRK2/3, LIMK1 and STK16,
all of which are not commonly associated with cancer, were the only other kinases significantly
inhibited by this compound. To the best of our knowledge, the kinase selectivity of 83 is superior
1
4,15
to those of CDK4/6 inhibitors previously reported.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 5. Kinomeꢀwide selectivity of 83 at 1 µM. Red, yellow and green colors indicate kinase
residual activity of < 10%, 10ꢀ20% and > 20%, respectively. Individual kinase residual activity
values are given in Table S2. The human kinome map is adapted with permission from Reaction
Biology Corp.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Anti-proliferative Activity of 83. The most selective CDK4/6 inhibitor 83 was further tested
against a panel of 12 cancerous and two nonꢀtransformed cell lines to examine whether the antiꢀ
proliferative effect was cellꢀtype specific (Table 5). Compound 83 potently inhibited the
proliferation of seven cell lines (MOLM13, NB4, M229, M249, M249R, A2780 and PCꢀ3) with
GI values between 0.038 and 0.681 ꢁM, and moderately restrained the growth of two cell lines
5
0
(
MCFꢀ7 and M238) with a GI value of 1.887 and 2.200 ꢁM, respectively. The remaining three
50
cancer cell lines were not sensitive to 83 with a GI50 value of > 10 ꢁM. In comparison,
palbociclib was more effective against the cancer cell lines except in the cases of melanoma
cancer cells M229, M249 and M247R. Both 83 and palbociclib exerted little toxic effects on
nonꢀtransformed human prostatic epithelial cell BPHꢀ1 and the normal lung fibroblast MRCꢀ5.
ACS Paragon Plus Environment
24

Page 25 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 5. Antiꢀproliferative activity of 83 against a panel of human cell lines
a
Growth inhibition, GI (ꢁM)
50
Origin
Breast
Cell lines
8
3
Palbociclib
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
MCFꢀ7
MDAꢀMBꢀ468
MOLM13
NB4
1.887 ± 0.130
> 10
0.557 ± 0.057
5.960 ± 2.400
0.062 ± 0.012
0.066 ± 0.012
1.220 ± 0.610
1.970 ± 0.260
1.990 ± 0.520
2.960 ± 0.220
0.032 ± 0.008
5.792 ± 0.801
4.150 ± 1.100
0.071 ± 0.009
7.442 ± 0.552
> 10
Leukemia
Melanoma
0.385 ± 0.098
0.681 ± 0.018
0.150 ± 0.007
2.200 ± 0.430
0.082 ± 0.002
0.038 ± 0.005
0.058 ± 0.017
> 10
M229
M238
M249
M249R
A2780
Duꢀ145
C4ꢀ2B
Ovarian
Prostate
> 10
PCꢀ3
0.150 ± 0.007
> 10
Nonꢀtransformed
BPHꢀ1
MRCꢀ5
> 10
a
Antiꢀproliferative activity was determined by 72 h resazurin or MTT assays. The data given
are derived from two replicates.
Cellular Mechanism of Action of 83. To evaluate whether the observed antiꢀproliferative
activity of 83 is derived from the inhibition of cellular CDK4/6, we examined the status of Rb
phosphorylation at the CDK4/6ꢀspecific site Ser780, cell cycle distribution, and apoptosis
induction. Western blot analysis showed that 83 inhibited phosphorylation of the Rb protein (at
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ser780) in MV4ꢀ11 cells in a doseꢀdependent manner upon 12 h and 24 h treatments (Figure 6).
In comparison, much faster and stronger inhibition of both total Rb and Rb phosphorylation was
observed upon treatment with palbociclib. This finding is consistent with a previous study, where
plabociclib was shown not only to decrease the protein levels of Rb and phosphorylated Rb but
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
9
also the transcription of Rb1 gene. As phosphorylated Rb protein is required for G1 to S
transition of the cell cycle, selective inhibition of Rb phosphorylation at CDK4/6ꢀspecific site
should cause G1 cell cycle arrest. As expected, a doseꢀdependent accumulation of cells in G1
phase was observed following 24 h treatment with 83 or palbociclib (Figure 7). This G1 phase
arrest was maintained with 10 x GI ꢁM (and higher concentrations, data not shown) for both
5
0
compounds signifying no offꢀtarget cellꢀcycle activity. The lack of an increase in the population
of subꢀG1 cells (Figure 7) in both cases is in agreement with the absence of noticeable apoptosis
(Figure 8). This was not unexpected as previous studies have demonstrated that CDK4/6
inhibitors induced either cytostasis or apoptosis depending on the type of cells, concentration of
4
0,41
drug and duration of treatement.
Taken together, these results indicate that 83 effectively
triggers G1 cell cycle arrest by inhibiting the CDK4/6ꢀmediated phosphorylation of Rb without
causing significant cell death in MV4ꢀ11 cells.
Figure 6. Western blot analysis of MV4ꢀ11 cancer cells following treatment with 83 or
palbociclib for 12 or 24 h. Both 83 and palbociclib (palb) inhibited the phosphorylation of Rb at
ACS Paragon Plus Environment
26

Page 27 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Ser780. βꢀActin was used as a loading control. The data are representative of two independent
experiments.
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 7. Cell cycle effects of 83 and palbociclib on MV4ꢀ11 cells. Cells were treated with
increasing concentrations of 83 or palbociclib (palb) for 24 h, harvested, fixed, and stained with
propidium iodide prior to flow cytometric analysis. The experiments were repeated twice.
Representative data were selected to generate this figure.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. Effects of 83 and palbociclib on the induction of apoptosis. Treatment with 83 or
palbociclib (palb) showed little effect on the induction of apoptosis in MV4ꢀ11 cells. The cells
were exposed to increasing concentrations of each compound for 24 h, and analyzed by annexin
V/propidium iodide double staining. The data are representative of two independent experiments.
Biopharmaceutical Profiling of 83. In light of its excellent in vitro biological profile, 83 was
further subjected to in vitro biopharmaceutical and in vivo pharmacokinetic evaluations. To
assess the safety and potential oral bioavailability of 83, the in vitro ADME properties of 83
were determined. CYP450 and hERG (human etherꢀaꢀgoꢀgo related gene) channel inhibition
screens were used to evaluate the potential of 83 towards drugꢀdrug interactions and cardiac
liability, respectively. As shown in Table 6, 83 exhibited a very good safety profile with IC >
5
0
2
5 µM against five common CYP450 isoforms, and with IC > 10 µM against hERG. 83 was
50
−
6
found to be highly permeable with apparent permeability coefficient (P ) of 10.1 × 10 cm/s,
app
suggesting high intestinal absorption. The efflux ratio (ER) of 83 was 0.4, demonstrating that 83
ACS Paragon Plus Environment
28

Page 29 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was not subjected to active efflux that is usually indicated by a value > 2. 83 has high
lipophilicity with a logD value of 4.4 (Table 6).
7.4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 6. Biopharmaceutical profiles of 83
a
Inhibition of CYP450 (IC , µM)
CYP3A4
CYP2D6
CYP1A
CYP2C9
CYP2C19
µM
>25
50
>25
>25
>25
>25
b
Inhibition of hERG (IC50)
>10
c
Partition coefficient
LogD7.4
pKa
4.4
d
Dissociation constant
3.98, 5.68
10.1
3.98
0.396
e
ꢀ6
Caco2 permeability
A−B Papp (10 cm/s)
ꢀ6
B−A Papp (10 cm/s)
efflux ratio
a
Human liver microsomes and NADPH in the presence of a cytochrome P450 isoformꢀspecific
probe substrate: midazolam for CYP3A4, dextromethorphan for CYP2D6, ethoxyresorufin for
b
CYP1A, tolbutamide for CYP2C9 and mephenytoin for CYP2C19. CHOꢀhERG cells were used
for the cardiac toxicity assay. The data is usually categorized into the following classification
bands: highly potent IC < 0.1 ꢁM; potent IC = 0.1 – 1 ꢁM; moderately potent IC = 1 – 10
5
0
50
50
c
d
ꢁ
M; weak or no inhibition IC > 10 ꢁM. Determined using a pHꢀmetric method. by Shake
50
e
Flask. Apparent permeability coefficient measured by Cacoꢀ2 assay. Efflux Ratio = (mean Papp
BꢀA / mean Papp AꢀB).
Pharmacokinetic Properties of 83. Encouraged by the in vitro DMPK profile of 83, singleꢀ
dose PK studies were then performed with male albino Wistar rats at 5 mg/kg by IV and 20
mg/kg by PO and with male BALB/c mice at 2 mg/kg by IV and 10 mg/kg by PO, and the results
ACS Paragon Plus Environment
29