Optimization of Pan-Pim Kinase Activity and Oral Bioavailability Leading to Diaminopyrazole (GDC-0339) for the Treatment of Multiple Myeloma

Article abstract of DOI:10.1021/acs.jmedchem.8b01857

Pim kinases have been targets of interest for a number of therapeutic areas. Evidence of durable single-agent efficacy in human clinical trials validated Pim kinase inhibition as a promising therapeutic approach for multiple myeloma patients. Here, we report the compound optimization leading to GDC-0339 (16), a potent, orally bioavailable, and well tolerated pan-Pim kinase inhibitor that proved efficacious in RPMI8226 and MM.1S human multiple myeloma xenograft mouse models and has been evaluated as an early development candidate.

Full text of DOI:10.1021/acs.jmedchem.8b01857

Subscriber access provided by EKU Libraries
Article
Optimization of Pan-Pim Kinase Activity and Oral Bioavailability Leading
to Diaminopyrazole (GDC-0339) for the Treatment of Multiple Myeloma
Xiaojing Wang, Wesley Peter Blackaby, Vivienne Allen, Grace Ka Yan Chan, Jae H Chang, Po-
Chang Chiang, Coura Diene, Jason Drummond, Steven Do, Eric Fan, Eric Harstad, Alastair
Hodges, Huiyong Hu, Wei Jia, William Kofie, Aleksandr Kolesnikov, Joseph p Lyssikatos, Justin
Q. Ly, Mizio Matteucci, John G. Moffat, Veerendra Munugalavadla, Jeremy M. Murray, David
Nash, Cameron Noland, Geoffrey Del Rosario, Leanne Ross, Craig Rouse, Andrew Sharpe,
Dionysos Slaga, Minghua Sun, Vickie Tsui, Heidi J. A. Wallweber, Shang-Fan Yu, and Allen Ebens
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01857 • Publication Date (Web): 04 Feb 2019
Downloaded from http://pubs.acs.org on February 5, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 50
Journal of Medicinal Chemistry
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 50
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
Optimization of Pan-Pim Kinase Activity and Oral
Bioavailability Leading to Diaminopyrazole
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
(GDC-0339) for the Treatment of Multiple
Myeloma
a,
b
b
a
a
Xiaojing Wang *, Wesley Blackaby , Vivienne Allen , Grace Ka Yan Chan , Jae H. Chang ,
a
b
a
a
a
a
Po-Chang Chiang , Coura Diène , Jason Drummond , Steven Do , Eric Fan , Eric Harstad ,
b
a
a
b
a
Alastair Hodges , Huiyong Hu , Wei Jia , William Kofie , Aleksandr Kolesnikov , Joseph P.
a
a
b
a
a
Lyssikatos , Justin Ly , Mizio Matteucci , John G. Moffat , Veerendra Munugalavadla ,
a
b
a
a
a
Jeremy Murray , David Nash , Cameron Noland , Geoff Del Rosario , Leanne Ross , Craig
b
b
a
a
a
a
Rouse , Andrew Sharpe , Dionysos Slaga , Minghua Sun , Vickie Tsui , Heidi Wallweber ,
a
a
Shang-Fan Yu , Allen J. Ebens
aGenentech, Inc., 1 DNA Way, South San Francisco, California 94080, United States.
bCharles River Discovery Research Services UK Limited, Chesterford Research Park, Saffron
Walden, Essex, CB10 1XL, United Kingdom.
ACS Paragon Plus Environment

Page 3 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
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
Pim kinases have been targets of interest for a number of therapeutic areas. Evidence of
durable single-agent efficacy in human clinical trials validated Pim kinase inhibition as a
promising therapeutic approach for multiple myeloma patients. Here, we report the compound
optimization leading to GDC-0339 (16), a potent, orally bioavailable, and well tolerated pan-
Pim kinase inhibitor that proved efficacious in RPMI8226 and MM.1S human multiple
myeloma xenograft mouse models and has been evaluated as an early development candidate.
Keywords:
Pim kinases; Kinase inhibitor; Oral bioavailability; Lead optimization; GDC-0339; multiple
myeloma
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 50
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
INTRODUCTION
Multiple myeloma is an incurable B-cell neoplasm characterized by the accumulation of
malignant plasma cells. Upon accumulation in the bone marrow, cancerous plasma cells
crowd out the healthy blood cells and cause bone lesions, increased calcium level, anemia,
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
and renal insufficiency. Multiple myeloma accounts for 10% of all hematological cancers,
2
and although rare, represents the second most common hematological cancer. Standard of
care therapies for the treatment of multiple myeloma, including DNA alkylators,
3
glycocorticoids, IMiDs and proteasome inhibitors , have evolved over the past half of a
4
century and significantly improved the 5-year survival from ~10% to 66% in 2014. Despite
effective therapies that can prolong patients’ lives, the quality of life for elderly and fragile
5
patients (median age at diagnosis is 70 years ) may be compromised due to the undesirable
side effects of current therapies. In addition, as health care improvements allow more people
to live into their seventies and beyond, higher incidence of multiple myeloma would be
expected, requiring safer and effective novel therapies.
The Proviral Integration site in Moloney murine leukemia virus (Pim) kinases are frequently
overexpressed in multiple myeloma, and inhibition of Pim kinase activity was found to be
beneficial for the treatment of multiple myeloma.6-9 The Pim proteins are a family of
constitutively active serine/threonine kinases (Pim-1, Pim-2 and Pim-3) that are highly
homologous. Besides multiple myeloma, Pim kinase inhibitors are potentially useful in
targeting other malignancies, including acute lymphoblastic leukemia (ALL), acute
myelogenous leukemia (AML), non-Hodgkin lymphoma (NHL) and triple-negative breast
cancer.10-14 Pim kinase inhibitors are therefore highly sought after as a new cancer therapy.
Numerous groups have reported the discovery of Pim kinase inhibitors,15-32 including previous
ACS Paragon Plus Environment

Page 5 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
work conducted in our laboratory (46 in Figure 1).33-35 To date, three Pim inhibitors have
entered Phase I clinical trials but none has been approved, i.e. SGI-1776,36-38 AZD120839,40
1
5
and LGH447 (Figure 1). LGH447 provided evidence of durable single-agent efficacy in
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
human clinical trials, validating Pim kinase inhibition as a promising therapeutic rationale for
7
multiple myeloma patients. In this article, we report the discovery and characterization of
GDC-0339, a pan-Pim kinase inhibitor clinical candidate for the treatment of multiple
myeloma.
Pim inhibitors tested in clinical trials
NH2
N
NH2
N
N
NH
O
N
N
F
NH
O
HN
O
F
F
N
N
S
O
F
F
F
1
(SGI-1776)
2 (AZD1208)
3 (LGH447)
Our previous work
HN
N
N
N
F
H N
O
2
N
N
N
N
N
H
O
NH
N
NH
N
N
N
N
F
N
HN
N
N
N
N
N
H
4
5
6
Figure 1. Selected Pim kinase inhibitors disclosed in the literature.
RESULTS AND DISCUSSION
Burger et al first disclosed pan-Pim kinase inhibitors such as 7 in Table 1, containing a
diaminopyridyl moiety (highlighted in red) that forms a hydrogen-bond with a catalytic Lys67
residue of Pim-1.16 Inspired by this work, we prepared compound 8 in which the 3-
aminopyridine (in black) was replaced by a bioisosteric 5-aminothiazole group. Compound 8
showed excellent biochemical potency (K : 0.01–0.05 nM), improved permeability (P A-B:
i
app
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 50
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
-
6
3
vs. 0.5  10 cm/s), and cellular potency (IC : 0.05 vs. 0.16 M), but decreased solubility
5
0
(28 vs. 934 M) and oral bioavailability (23% vs. 54%) compared to 7. To mitigate potential
liabilities such as hERG inhibition, diaminopyrazole moiety was used in place of the
diaminopyridine. This change, combined with expansion of the piperidine to the
homopiperidine for optimal placement of the primary amine provided a prototype molecule 9.
Compound 9 displayed a nice combination of the best properties of 7 and 8 in terms of
biochemical potency, cellular potency, and solubility. In addition, the potency of 9 against the
hERG potassium channel is 100-fold lower than that of 8 (10.1 and 0.11 M respectively,
Table S2 in supporting information). However, 9 suffered from low permeability and high
transporter efflux ratio in MDCK epithelial cell lines, resulting in low oral bioavailability in
rats.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Identifying hits using literature aided drug design.
ID Structure
In vitro
In vivo In silico
a
Pappd Rat
g
K (nM)
MM.1S
Sol.
c_pKa
i
c
e
h
Pim-1
Pim-2
Pim-3
IC50
(M) A-B CL
cgLogD
b
f
i
(M)
B-A F%
ER
tPSA
N
NH2
0.03
0.2
N
0.5
15
6
8.4
1.5
110
NH
7
8
N
0
.16
934
28
O
0
.06
54%
F
H2N
1
3
N
NH
NH2
F
N
0
0
0
.01
.05
.02
3
9
3
8.4
1.6
110
16
O
N
0.05
0.08
2
3%
S
H2N
F
NH2
N
N
N
0.02
.1
0.007
0.6
2523 11
18
9.3
1.2
115
9
NH
12
F
F
0
O
N
8
%
S
H2N
ACS Paragon Plus Environment

Page 7 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
aBiochemical assays against three Pim isoforms; geometric mean values are shown here; 95%
b
confidence intervals where n > 3 are in Table S1. Cellular proliferation assay; geometric
mean values are shown here; 95% confidence intervals where n > 3 are in Table S1.
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
c
d
Measured thermodynamic solubility at pH 6.5. P : measured permeability using Madin-
app
-6
Darby Canine Kidney (MDCK) epithelial cell lines, A to B or B to A (10 cm/s); ER: efflux
e
ratio calculated as Papp (B to A) divided by Papp (A to B). Total clearance in rat in vivo
(
(
(
mL/min/kg); compound was dosed at 1 mg/kg in a solution of DMSO/Cremphor/water
f
5/5/90) for 7, 8, and 9. Rat bioavailability was calculated using (AUC ÷AUC ) x
oral
iv
Dose ÷Dose ) x 100; compound was dosed at 5 mg/kg in a solution of PEG400/water for 7
iv
oral
and 8, a suspension of MCT in water for 9; red denotes decreased bioavailability compared to
g
h
9
. Calculated logarithm of acid dissociation constant of the most basic amine. Calculated
i
Genentech logarithm of distribution-coefficient at pH 7.4. Calculated topological polar
surface area.
A co-crystal structure of compound 9 bound to Pim-1 at 2.39 Å resolution provided further
insight into the pharmacophore (Figure 2). Besides van der Waals interactions, several key
hydrogen-bond interactions were identified and represented by dashed lines in Figure 2.
Consistent with our initial design hypothesis, the pyrazole nitrogen forms a hydrogen-bond
with Lys67. The carbonyl oxygen of the amide was engaged in a network of hydrogen-bonds
with Glu89 and Asp186 through stable water molecules. The aminothiazole is within
hydrogen-bond distance to the backbone carbonyl oxygen of Glu121. Due to the resolution,
the 4-aminoazepane moiety was not explicitly defined. We speculated that the primary amine
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 50
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
would make salt bridges with the acidic residues (Asp128 and Glu171) and the surrounding
water molecules.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. 2.39 Å crystal structure of compound 9 (PDB code: 5V80, authors will release the
atomic coordinates and experimental data upon article publication) bound to Pim-1. Hydrogen
bond interactions are depicted as black dashed line.
In rat pharmacokinetics (PK) studies, compound 9 suffered from low oral bioavailability. At
pH 7.4, a total of six hydrogen bond donors (HBD) are displayed in 9 which may contribute to
4
1
the low permeability and high efflux in the MDCK cell lines. Therefore, we set out to reduce
the total number of HBD to improve the oral bioavailability. Compound 10, with a 1,4-
diazepane replacing the 4-aminoazepane (Table 2), suffered from a drastic loss in both
biochemical (10–50-fold) and cellular potency (163-fold). While the basic primary amine of 9
is well positioned to form multiple salt-bridge interactions to the acidic patch formed by
Asp128 and Glu171 residues, the secondary amine of 10 may not be close enough to maintain
similar beneficial interactions, resulting in the observed potency loss for 10. However, with
one less HBD and lower basicity of the amine (c_pK : 7.5 for 10 vs. 9.3 for 9), 10
a
ACS Paragon Plus Environment

Page 9 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
demonstrated significantly improved oral bioavailability (F%: 67% vs. 8%) commensurate
with its higher kinetic solubility (127 vs. 55 M) and permeability (Papp A-B: 5.9 vs. 0.6  10-
6
cm/s) relative to 9. Compound 11 with an alcohol in place of the primary amine in 9, also
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
suffered a loss of potency (4–30-fold in biochemical and 26-fold in cellular assays). The
-6
modest increase in permeability (Papp A-B: 4.6 vs. 0.6  10 cm/s) was offset by the loss of
solubility (< 1 vs. 55 M), which ultimately resulted in a modest gain in oral bioavailability
(F%: 38% vs. 8%) compared to 9. Based on the co-crystal structure of 9 with Pim-1, we
observed that the sulfur atom of the aminothiazole moiety is in close proximity (3.98 Å) to the
backbone carbonyl oxygen of Glu121, providing a potential opportunity for an OS -hole
interaction.42 We therefore made 12 without a NH group on the thiazole, hoping that the
2
thiazole sulfur would be closer to the carbonyl oxygen of Glu121. Unfortunately, the modest
-6
gain in permeability (Papp A-B: 1.6 vs. 0.6  10 cm/s) and oral bioavailability (F%: 32% vs.
8
%) were not adequate to offset the marked potency loss comparing 12 to 9 (30–80-fold in
biochemical and > 313-fold in cellular assays). While decreasing the number of HBD resulted
in the loss of potency, the enhanced permeability and oral bioavailability encouraged us to
continue the strategy of optimizing physiochemical properties to improve PK profile.
Table 2. Reducing the number of hydrogen-bond donors to improve permeability and
oral bioavailability.
R2
N
N
NH
F
O
N
R1
S
F
ID R1
R2
In vitro
Ki
(
In vivo In silico
g
MM.1S
Kin.
Pappd Rat
NH+OH
nM)a
IC (M) Sol.
b
c_pKa
h
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 50
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
Pim-1
Pim-2
Pim-3
0.02
(M)c A-B CLe
f
B-A F%
ER
NH2
0.6
12
11
18
9
-NH2
6
9.3
0.1
0.08
13
55
N
N
8
%
0
0
5
0
.007
.2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5.9
21
3.6
4.6
4
0.9
1.6
17.6
11
NH
10 -NH2
9
5
7.5
127
< 1
50
67%
7
.2
OH
0.09
2
1
1i -NH2
4
3.2
2.1
N
N
3
8%
0
.2
NH2
0.6
12 -H
40
2%
4
9.3
8
0
.1
.4
>25
3
a,b,c,d
see footnotes from Table 1; red denotes significant erosion on potency compared to 9;
e
green denotes improvement compared to 9. Total clearance in rat in vivo (mL/min/kg);
compound was dosed at 1 mg/kg in a solution of DMSO/PEG400/water for 10, 11, and
f
PEG400/water (40/60) for 12. Rat bioavailability calculated using (AUC ÷AUC ) x
oral
iv
(Dose ÷Dose ) x 100; compound was dosed at 5 mg/kg in a suspension of MCT in water for
iv
oral
g
1
0, 11, and 12; green denotes improvement over 9. Calculated total number of active protons
h
at pH 7.4. Calculated logarithm of acid dissociation constant of the most basic amine.
iCompound 11 is a eutomer with unknown stereochemistry.
In parallel, we investigated the effects of reduced basicity of the primary amine on
permeability and oral bioavailability. Direct substitution of a trifluoroethyl (13) or
difluoroethyl (14) group on the primary amine significantly reduced the c_pK , but
a
unfortunately was not well tolerated in terms of biochemical potency (Table 3). Difluoro
substitutions at the -position of the primary amine as exemplified by 15 maintained
biochemical potency but had weak cellular potency, most likely a result of the increased
lipophilicity and plasma protein binding (Table S2). Monofluoro substitution at the -position
ACS Paragon Plus Environment

Page 11 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of the primary amine in 16 or 17 was well tolerated with a concurrent improvement in
permeability and oral bioavailability. Given these encouraging results, all other stereoisomers
of 16 and 17 were characterized but overall, none of them was superior to 16 and 17. When
compared to 17, 16 showed better Pim-2 binding potency (4-fold), better oral bioavailability,
and was easier to synthesize. Difluoro substitution at the -position of the primary amine (18
and 19) was also well tolerated in the biochemical assay. Compound 19 displayed a larger
cellular shift than 18, again potentially due to the higher lipophilicity and plasma protein
binding of 19 (Table S2). Analog 18 had improved permeability (Papp A-B: 3.4 vs. 0.6  10-6
cm/s) and oral bioavailability (F%: 71% vs. 8%) compared to 9. Mono-substitution on either
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

 or position of the primary amine such as in 20 – 23 maintained excellent potency,
however only modestly reduced the calculated c_pK (< 0.7) of the most basic amine and
a
consequently had poor permeability, high efflux and low oral bioavailability compared to 9.
In summary, there seemed to be a trend that permeability and oral bioavailability can be
improved by adjusting the basicity of the most basic primary amino group in the azepane
subseries. Lastly, compounds such as 24 and 25, with a heptane ring attached to the central
pyrazole ring instead of an azepane, maintained good potency and had a modest improvement
in permeability or oral bioavailability, however both required complex synthesis.
Table 3. Reducing the basicity of the primary amine to improve permeability and oral
bioavailability.
R
N
N
NH
F
O
N
S
H N
2
F
ID R
In vitro
In vivo
In silico
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 50
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
Ki
(
Pim-1
Pim-2
Pim-3
MM.1S
IC (M)
Kin.
Sol.
(M)
Pappd Rat
A-B CLe
B-A F%
ER
c_pKah
nM)a
b
5
0
c
f
NH2
0.02
0.6
12
11
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
9
0
0
.1
.007
0.08
55
9.3
N
N
8
%
18
HN
CF3 0.2
1
1
1
1
1
1
1
2
2
2
2
2
3
4
5
6
7
8
9
0
1
2
3
4
9
ND
8
1
ND
ND
ND
4.4
6.0
6.2
7.8
7.8
8.4
8.5
8.6
8.7
8.7
8.8
9.1
0
CF2H 0.08
6
.02
HN
N
1
ND
0.005
0.04
NH2
0.4
28.1
11
27
91%
F
F
0.3
.1
0.5
2.8
67
N
N
0
NH2
F
0.03
0.1
0.02
7
14.0
2
8
3
0.07
0.1
3%
0
0
0
.08
.4
.009
6.5
13.1
2
12
24%
108
81
NH2
0.005
.04
0.009
.003
ND
.01
3.4
12.6
3.8
14.5
26
71%
0
0.04
0.9
N
N
F
F
NH2
0
164
52
15.2 ND
1.1
F
F
0
NH2
O
0.05
0.2
0.19
0.03
0.2
0.9
12.1
13
5
1
0.05
0.03
0.07
0.008
0.2
N
N
6%
NH2
0.1
4.3
32
15
3%
OH
100
112
62
0
0
0
.01
.01
.05
1.4
10.6
7.7
n/a
9.4
-
3.4
n/a
-
25
5
%
0.01
NH2
O
0
.007
2
N
0.04
3%
ND
0
.02
Abs
NH2
F
0.03
0.3
0.01
88
ACS Paragon Plus Environment

Page 13 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH2
0.006
ND
2.8
5.4
3.2
3
24%
25
O
0.07
16
9.1
0
.005
a,b,c,d
See footnotes from Table 1; ND: not determined, n/a: not available due to low mass
e
recovery. Total clearance in rat in vivo (mL/min/kg); compound was dosed at 1 mg/kg in a
solution of DMSO/PEG400/water for 1518, 2123, and 25, PEG400/water (40/60) for 20.
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
f
Rat bioavailability calculated using (AUC ÷AUC ) x (Dose ÷Dose ) x 100; compound
oral
iv
iv
oral
was dosed at 5 mg/kg in a suspension of MCT in water for 1518, 20–23, and 25; green/red
h
denotes better/worse than 9. Calculated logarithm of acid dissociation constant of the most
basic amine.
Based on their overall profiles balancing potency, PK and ease of synthesis, we considered
compounds 16 and 18 for additional characterization. One differentiating property between 16
and 18 was time dependent inhibition (TDI) activity of CYP3A4. Since Pim inhibitors have a
high potential to be combined with other therapeutic agents for improved efficacy, drug-drug
interaction (DDI) as a result of TDI is highly undesired.43 As shown in Table 4, both
compounds exhibited reversible inhibition of CYP3A4 (IC : 3.8–7.5 M) using either
5
0
midazolam (M) or testosterone (T) as the probe substrate. While 16 exhibited a moderate TDI
liability with low to moderate DDI potential, the TDI risk associated with 18 was much higher
and similar to the positive control mifepristone (kinact/K ratio: 64). We previously reported our
I
efforts in removing the TDI liability of 18 by methyl substitution at the -position of the basic
4
4
amine as in 19. Unfortunately, cellular activity of 19 was poor, likely due to high plasma
protein binding. In addition, compound 16 compared favorably over 18 with higher selectivity
in the kinase panel (Table S7). Based on all these, 16 was selected for further profiling.
Table 4. CYP3A4 reversible and time dependent inhibition (TDI) of 16 and 18.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
6
18
Rev.
M
T
7.5±2.2 3.8±2.5
6.3±3.3 4.9±2.9
IC ±SD
5
0
a
(
M)
b
TDI
K (M)
I
1.12
1.16
0.061
52
-1
kinact (min ) 0.027
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
Ratio
24
a
CYP3A4 reversible inhibition assay using either midazolam (M) or testosterone (T) as the
probe substrate; IC was reported as a mean +/- standard deviation with n=10 for 16 and n=7
5
0
b
for 18. CYP3A4 time dependent inhibition using testosterone as a probe; K is the inhibitor
I
concentration that supports half the maximal rate of inactivation; kinact is the maximal rate of
enzyme inactivation; ratio calculated using 1000 x kinact/K (mL/min/mol).
I
A co-crystal structure of 16 and Pim-1 was obtained to confirm the binding mode. Similar to 9,
compound 16 was well positioned in the Pim-1 adenosine triphosphate (ATP) binding pocket
with multiple van der Waals and polar-polar interactions. By overlaying both co-crystal
structures, it was clear that the aromatic rings were perfectly aligned. In the co-crystal
structure with 16, the aminofluoroazepane moiety was well resolved with clear salt bridge
interactions from the primary amine to Asp128 and Glu171.
ACS Paragon Plus Environment

Page 15 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 3. 1.71 Å crystal structure of compound 16 (PDB code: 6NO9, authors will release the
atomic coordinates and experimental data upon article publication) bound to Pim-1. Hydrogen
bond interactions are depicted as black dashed line.
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
Kinase selectivity of 16 was obtained against a panel of 277 kinases (Table S7). When tested
at 0.1 M (500–5000-fold above Pim K ), 16 showed higher than 50% binding to only 12
i
kinases. K values for these 12 kinases were determined and only 6 kinases had a lower than
i
5
00-fold K ratio over Pim-2 (Table 5). Among these, we ran a dose-response in the cellular
i
assays for mitogen-activated protein kinase kinase kinase kinase 4 (MAP4K4) and glycogen
synthase kinase 3 (GSK3. The IC of 16 in each cellular assay were in the micromolar
5
0
range and were at least 57-fold higher than that in MM.1S assay. We also tested 16 against a
non-kinase selectivity panel, including G-protein coupled receptors (GPCRs), ion channels,
and transporters. When 16 was tested at a concentration of 10 M, 6 out of 39 targets showed
higher than 50% binding. In functional cell-based assays for these 6 targets, only 3 of them
yielded a measurable IC and these were in the micromolar range. The IC against dopamine
5
0
50
(
DOP) receptor was 0.5 M; however, this may not have any physiological consequence since
6 is not brain penetrant. Compound 16 inhibited hERG potassium channel in a patch-clamp
PC) study with an IC of 2.7 M, which provided a 70-fold window to the unbound
1
(
5
0
maximum concentration (0.039 M) at the projected human efficacious dose equivalent to a
4
5
tumor growth inhibition of 90% in the RPMI8226 xenograft model (Figure 6). In limited 7-
day dose ranging studies in rats and monkeys, 16 was well tolerated at doses up to 100
mg/kg/day, which was at or above the projected human therapeutic plasma exposure. In all,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
these studies illustrated a favorable safety profile that supported more comprehensive safety
evaluation to enable clinical evaluation of 16.
Table 5. Selectivity and in vitro safety profile of 16.
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
Assay a
Result
Kinase
selectivity
Binding% > 50% @ 0.1 M
12 out of 277 kinases
DMPK (24x), PASK (43x), MAP4K4
(48x), GSK3 (52x), GSK3 (133x),
MNK1 (152x)
K ratio < 500 (kinase/Pim-2)
i
Cellular IC (uM)
MAP4K4: 5.7 [57x]
50
[fold over IC in MM.1S]
GSK3: 23 [230x]
5
0
Non-
kinase
selectivity
Binding% > 50% @ 10 M
6 out of 39 enzymes
b
alpha 1a: 2.5
c
Functional IC (M)
Na PC tonic/phasic: 11.5/8.8
5
0
d
b
e
DOP : 0.5; 5HT1a: 4.9 /2.5
f
hERG PC IC (M)
2.7
5
0
a
b
See the supporting information for assay methods and details. Antagonist assay for alpha 1a
c
adrenergic receptor (alpha 1a) or serotonin 1a receptor (5HT1a). Sodium channel patch clamp
d
e
f
assay. DOP, dopamine receptor. Agonist assay for 5HT1a. hERG potassium channel patch
clamp assay.
The Absorption, Distribution, Metabolism, and Excretion (ADME) properties of 16 across
species are listed in Table 6. Predicated hepatic stability (assuming the same unbound
fractions in blood and microsomes) in liver microsomes and hepatocytes was low to moderate
and relatively similar between the two assays, suggesting that the main route of metabolism
was phase 1. The in vivo clearance varied and was the lowest in rat and highest in mouse. Oral
bioavailability was moderate in rat and cyno but low in mouse. Based on these preclinical data,
both the clearance and oral bioavailability were anticipated to be moderate in human.
Table 6. Preclinical ADME profiling of 16.
Species
In vitro
In vivo
ACS Paragon Plus Environment

Page 17 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CLc F%d
Vssc T1/2c
LM
CLhep CLhep
76
39
21
34
Hep
a
b
e
f
mouse
rat
54
37
135 11 – 34 6.8 0.9
e
f
8
33 – 28 0.9 1.8
--
-- --
35 – 60 5.0 1.8
-- -- --
dog
cyno
21
24
10
--
34
--
g
f
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
human 11
a
b
Projected hepatic clearance using liver microsomes (mL/min/kg). Projected hepatic
c
clearance using hepatocytes (mL/min/kg). Total clearance in vivo (CL, mL/min/kg); volume
of distribution at steady state (V , L/kg); half life (T , hr); compound was dosed at a 1 mg/kg
ss
1/2
dose in mouse (n = 2) or rat (n = 2) in a solution of DMSO/PEG400/water and at a 0.5 mg/kg
d
dose in cynomolgus (cyno) monkey (n = 2) in a solution of PEG400/water (40/60). F% is oral
e
bioavailability with 16 dosed in a suspension of MCT. Amorphous 16 in a 5 mg/kg dose (n =
f
g
2
). Crystalline 16 in a 100 mg/kg dose (n = 3). Amorphous 16 in a 1 mg/kg dose (n = 2).
The physiological downstream signaling of Pim kinases is mediated through the
phosphorylation of a large number of substrates (Figure 4a), including regulators of protein
translation (4EBP1, S6 ribosomal protein) and a regulator of cellular apoptosis (BAD).
Western blot analysis of MM.1S myeloma cell line treated with a dose range of compound 16
revealed a constellation of Pim down-stream signaling events consistent with inhibition of
Pim kinases (Figure 4b). These data demonstrated the dose-dependent modulation of
phospho-BAD (pBAD), phospho-4EBP1 (p4EBP1) and phospho-S6 ribosomal protein (pS6).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a)
Figure 4. Compound 16 Modulates known Pim substrates. Protein level by western blot
analysis after 4 hour treatment with DMSO or increasing concentrations of 16 in MM.1S cells.
b)Schematic illustration of Pim-kinase signaling pathway.
Given that Pim kinases share common substrates with the PI3K-AKT-mTOR signaling
pathway,46,47 a dose-response experiment combining both Pim inhibitor 16 and
4
8
phosphoinositide 3 kinase (PI3K) inhibitor GDC-0941 was conducted. Compound 16
inhibited cell viability at a maximum of -60% (the bottom of the dose-response curve) in
contrast to GDC-0941 at -100%. This was consistent with 16 being cytostatic while GDC-
0
941 being cytotoxic. The relative IC of the combination treatment improved by 3-fold
50
compared to each single agent, indicating a potential synergistic combination.
ACS Paragon Plus Environment

Page 19 of 50
Journal of Medicinal Chemistry
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 5. Compound 16 is active in multiple myeloma cancer cells both as a single agent and
in combination with PI3K inhibitor GDC-0941.
Compound 16 was then evaluated in vivo using multiple myeloma xenograft mouse models.
In the RPMI8226 mouse model, 16 was orally administered with doses ranging from 1 to 300
mg/kg once per day for 21 days. Mice were well tolerated with body weight change less than
1
0%. Both clear dose-response and tumor regression were observed with 16 dosed as a single
agent (Figure 6a). In a separate experiment, pharmacodynamic (PD) markers pS6 and pBAD
were also shown to decrease with increasing doses of 16 (Figure 6b,c), consistent with the
western blot data.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 50
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 6. Single agent activity of 16 was observed in a RPMI8226 human multiple myeloma
xenograft mouse model. a) Effect of 16 on tumor growth. Female C.B-17 SCID mice bearing
subcutaneous tumors were administered with escalating doses of 16 or vehicle (0.5%
methylcellulose/0.2% Tween-80) orally once daily for 21 days. Changes in tumor volumes
over time for each treatment group are depicted as cubic spline fits generated via linear mixed
effects analysis of log-transformed volumes. b,c) Effect of 16 on Pharmacodynamic (PD)
markers. Tumors (n = 5/group) were excised from animals at indicated time after a single dose
of 16 and processed for analysis of Pim kinase pathway markers as described in the
supplemental method section. Levels of pS6 (Ser240/244) and pBAD (Ser112) were
normalized to the matched total protein and shown as mean (±SEM) for each treatment group.
ACS Paragon Plus Environment

Page 21 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A second multiple myeloma model, MM.1S, was also evaluated dosing mice with 16 orally
once per day. Similar to the RPMI8226 model, dose-dependent tumor growth inhibition was
observed albeit with higher doses required for the same extent of inhibition. All mice were
well tolerated with body weight change less than 15%. Tumor regression was observed at 450
mg/kg. However, given the significant weight loss at 450 mg/kg, data were not reported here.
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 7. Single agent activity of 16 was observed in a MM.1S human multiple myeloma
xenograft mouse model. Female C.B-17 SCID mice bearing subcutaneous tumors were
administered escalating doses of 16 or vehicle (0.5% methylcellulose/0.2% Tween-80) orally
once daily for 21 days. Changes in tumor volumes over time for each treatment group are
depicted as cubic spline fits generated via linear mixed effects analysis of log-transformed
volumes.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 50
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
To study the combination potential of Pim and PI3K inhibitors in vivo, we tested prototype
compound 8 with GDC-0941 in xenograft mouse models early on before 16 was discovered.
In the RPMI8226 xenograft mouse model, 8 and GDC-0941 combined well rendering a better
tumor growth inhibition than each single agent alone (Figure 8a). Such combination effect
was also evidenced by the tumor growth inhibition at the end of the study across 4 different
multiple myeloma xenograft mouse models (Figure 8b). When used as single agents, Pim
inhibitor 8 achieved slightly better tumor growth inhibition than PI3K inhibitor GDC-0941 in
all models except NCI-H929. However, the combo group achieved better tumor growth
inhibition than each single agent alone in all models.
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. Combination effect of 8 with a PI3K inhibitor (GDC-0941) was observed in human
multiple myeloma xenograft models. Female C.B-17 SCID or SCID-beige mice bearing
subcutaneous tumors were administered daily oral doses of vehicle (60% PEG400/40% water),
8
in 50 mg/kg, GDC-0941 in 75 mg/kg or the combination of 8 and GDC-0941 for 21 days.
a)Changes in tumor volumes over time for each treatment group in a RPMI8226 model are
ACS Paragon Plus Environment

Page 23 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
depicted as cubic spline fits generated via linear mixed effects analysis of log-transformed
b)
volumes. Tumor growth inhibition (TGI) at the end of the study is shown for each treatment
group in a RPMI8226 (Fig. 8a), MM.1S, NCI-H929 and OPM-2 model. TGI was calculated as
percent area under the fitted tumor volume-time curve (AUC) per day in relation to the
vehicle, using the formula: %TGI = 100 x [1 – (AUCtreatment/day ÷ AUCvehicle/day)].
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
The synthesis of building block 30 was accomplished in 4 steps and 16% overall yield
(Scheme 1). Homoallyl bromide and ammonium in methanol were heated to yield precursor
2
7. Homoallyl bromide and 27 were then heated under weakly basic condition to yield the
secondary amine which was then protected with a tert-butyloxycarbonyl (Boc) group to
provide 28. Cyclization of 28 using Hoveyda-Grubbs catalyst yielded the Boc protected
azepine 29. Subsequent deprotection under acidic condition yielded the HCl salt of building
block 30. This reaction sequence was carried out on a multi-gram scale.
Scheme 1. Synthesis of building block 30.
i
ii
N
Br
NH2
Boc
28
26
27
iii
iv
N
N
H
Cl
H
Boc
30
29
Reagents and conditions: (i) 12 M NH in MeOH, 90 C, 16 hr, 88%; (ii) (a) 26 in THF,
3
NaHCO , 60 C, 2 d; (b) Boc O, (Et) N, THF, 0 C to rt, 16 hr, 35% for two steps; (iii)
3
2
3
Hoveyda-Grubbs catalyst, DCM, rt, 2 d, 53%. (iv) 4 N HCl in dioxane, MeOH, rt, 4 hr, 95%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 50
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 azepine 30 in hand, the building block 38 was synthesized in 7 steps and 44% yield prior
to the SFC chiral separation (Scheme 2). We could not find examples in the literature of
straightforward way to synthesize 4,5-diaminopyrazole substituted at the N1 position so we
developed a facile regioselective synthesis. The readily commercially available 1-methyl-3-
nitropyrazole 31 was treated with lithium hexamethyldisilazane (LiHMDS) and
hexachloroethane at low temperature to yield the desired 5-chloro-1-methyl-4-nitropyrazole
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. Subsequent nucleophilic aromatic substitution (S Ar) reaction of 32 and 30, epoxidation
N
of 33, and azide addition of 34 were all accomplished in excellent yields to generate 35.
®
Deoxyfluorination using Deoxo-Fluor led to compound 36 in a good yield, with the fluorine
group in a trans orientation with respect to the azide group. Isolation of the trans diastereomer
was surprising since typically, deoxyfluorination leads to an inversion of the alcohol
stereocenter via a S 2 reaction mechanism. However, in our case, we believe azide group
N
participated and intervened with the transition state and consequently led to the double
inversion product 36. A similar neighboring group participation has been reported
previously.49,50 Reduction of the azide group by Staudinger reaction, followed by Boc
protection, and supercritical fluid chromatography (SFC) chiral separation yielded building
block 38 as a single stereoisomer.
Scheme 2. Synthesis of building block 38.
ACS Paragon Plus Environment

Page 25 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
N
NO2
3
1
i
N3
O
Rel
OH
Cl
NO2
ii
iii
iv
N
N
N
N
N
N
N
N
N
N
N
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
NO2
NO2
NO2
3
2
33
34
35
v
NHBoc
F
NH2
N3
Rel
N
Rel
N
Abs
N
F
F
vii
vi
N
N
N
N
N
N
NO2
NO2
NO2
38
37
36
Reagents and conditions: (i) Cl CCCl , LiHMDS, THF, -70 C, 4 hr, 68%; (ii) 30, KF, DMSO,
3
3
70 C, 16 hr, 98%; (iii) mCPBA, DCM, rt, 3 hr, 98%. (iv) NaN , NH Cl, MeOH/H O, 70 C,
3
4
2
®
1
6 hr, 100%; (v) Deoxo-Fluor , DCM, rt, 16 hr, 82%; (vi) Ph P, THF/H O, 60 C, 3 hr, 88%;
3
2
(
vii) (a) (Boc) O, DIPEA, DCM, rt, 16 hr, 94%; (b) SFC chiral separation.
2
The thiazole-containing building block 44 was prepared in 5 steps and 23% yield (Scheme 3).
Ethyl (hydroxyimino)cyanoacetate 39 is readily commercially available and was reduced to
aminocyanoacetate which was coupled with 2,5-difluorobenzoyl chloride in situ to yield
intermediate 41. Treatment of 41 with Lawesson’s reagent, followed by Boc protection, and
hydrolysis of the ester furnished building block 44.
Scheme 3. Synthesis of building block 44.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 50
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
F
O
O
O
i
ii
H
N
NH2
N
EtO
OH
EtO
EtO
41
CN
CN
40
CN
O
F
39
iii
O
O
F
F
F
F
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
HO
BocHN
N
S
v
EtO
R
N
S
N
H
44
42 R = H
43 R = Boc
iv
Reagents and conditions: (i) aq. NaHCO , Na S O , H O, 4 C to rt, 16 hr; (ii) 2,5-
3
2
2
4
2
difluorobenzoyl chloride, NaHCO , DCM, 0 C, 50 min, 59% for two steps; (iii) Lawesson’s
3
reagent, pyridine, reflux, 16 hr, 98%. (iv) (Boc) O, DMAP, CH CN, rt, 12 hr, 41%; (v) (a)
2
3
LiOH•H O, MeOH/H O, 50 C, 14 hr; (b) 2 M HCl, 0 C, 96%.
2
2
Compound 16 was made in 3 steps and 32% yield starting from the two key building blocks
8 and 44 (Scheme 4). Iron-mediated reduction of the nitro group in 38 yielded
3
aminopyrazole 45 which was then coupled with carboxylic acid 44 to yield 46. Acid-
promoted deprotection of the Boc group in 46, afforded 16 in excellent yield. It is worth
noting that a single crystal structure of 16 was obtained, confirming the stereochemistry
(
Figure S1 and Table S6).
Scheme 4. Synthesis of final compound 16.
Abs
H
N
R
NHBoc
NHBoc
F
N
Abs
N
Abs
N
N
N
F
F
i
ii
NH
N
F
N
N
S
O
N
N
NO2
NH2
45
R
N
H
F
iii
3
8
46 R = Boc
6 R = H
1
ACS Paragon Plus Environment

Page 27 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Reagents and conditions: (i) NH Cl, Fe powder, EtOH/H O, 99 C, 2.75 hr, 91%; (ii) 44,
4
2
DIPEA, PyBOP, DCM, rt, 48 hr, 39.8%; (iii) 4 N HCl in 1,4-dioxane, MeOH, rt, 20 hr, 89%.
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
CONCLUSIONS
In summary, a novel scaffold containing a diaminopyrazole moiety was described as pan-Pim
inhibitors. The development of a convenient and practical synthesis enabled the SAR
development and scale-up synthesis for both efficacy and safety studies. Through lead
optimization of the PK properties balanced with potency, selectivity, and ease of synthesis, 16
(GDC-0339) was identified as an early development candidate. Compound 16 is a potent and
efficacious pan-Pim inhibitor, orally bioavailable, and well tolerated in safety studies.
Compound 16 combined well with PI3K inhibition in vitro by decreasing IC and likely also
5
0
in vivo by enhancing tumor growth inhibition. Looking forward, Pim inhibition may be also
5
1
useful in combination with other pathway targets such as USP7, or applied to other disease
12
indications such as T-cell acute lymphoblastic leukemia subset and triple-negative breast
cancer. 13,14,52
EXPERIMENTAL SECTION
All chemicals (including 31 and 39) and solvents were used directly as received from
commercial suppliers. Syntheses of compounds 1 – 716,17,34-36,40,53 and 18 – 19 were reported
previously. Syntheses and characterizations of compounds 8 – 15, 17, and 20 – 25 are in
44
1
supporting information. H NMR spectra were recorded on Bruker Avance 400 or 500
spectrometers. Chemical shifts are expressed in δ ppm referenced to an internal standard,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 50
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
tetramethylsilane (δ = 0 ppm). Abbreviations used in describing peak signals are: br = broad
signal, s = singlet, d = doublet, dd = doublet of doublets, t = triplet, q = quartet, m = multiplet.
All final compounds were purified to have purity higher than 95% by reverse phase high
performance liquid chromatography (HPLC), supercritical fluid chromatography (SFC) or
normal phase silica gel chromatography flash chromatography. The purity was assessed by
reverse phase HPLC with an isocratic gradient of 5-95% acetonitrile in water (with either acid
or base modifier) and monitored by ultraviolet diode array detection at wavelength 254 nm.
Optical rotation was recorded using Autopol VI automatic polarimeter from Rudolph
Research Analytical. Low-resolution mass spectra (LCMS) were recorded in a liquid
chromatography-mass spectrometer in electrospray positive (ES+) mode. High-resolution
mass spectra (HRMS) experiments were performed on a Dionex LC Ultimate 3000 coupled
with a Thermo Scientific Q Exactive Orbitrap mass spectrometer using ESI as the ionization
source and a Phenomenex XB-C18, 1.7 mm, 50 mm x 2.1 mm column with a 0.7 mL/min
flow rate at 40 C for liquid chromatography (LC) separation. Solvent A was 0.1% formic
acid in water and solvent B was 0.1% formic acid in acetonitrile. The gradient consisted of 2-
98% solvent B over 7 min and held at 98% B for 1.5 min following equilibration for 1.0 min.
The LC was monitored by UV absorbance at 220 and 254 nm. MS full scans with 35000
resolution were applied to all experiments.
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
5
-Amino-N-(5-((4R,5R)-4-amino-5-fluoroazepan-1-yl)-1-methyl-1H-pyrazol-4-yl)-2-(2,6-
difluorophenyl)thiazole-4-carboxamide (16, GDC-0339). tert-butyl (4R,5R)-1-(4-(5-tert-
butoxycarbonylamino-2-(2,6-difluorophenyl)thiazole-4-carboxamido)-1-methyl-1H-pyrazol-
5-yl)-5-fluoroazepan-4-ylcarbamate (10.11 g, 15.2 mmol) in 4 N HCl in dioxane (200 mL)
ACS Paragon Plus Environment

Page 29 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and methanol (200 mL) was stirred at room temperature for 20 hr. Evaporation under reduced
pressure afforded a pale brown solid which was dissolved in 50% methanol in
dichloromethane and added to an SCX cartridge (strong cation exchange chromatography).
After washing with dichloromethane and methanol, elution with 1 N ammonia in methanol
and evaporation of the eluent under reduced pressure afforded 5-amino-N-(5-((4R,5R)-4-
amino-5-fluoroazepan-1-yl)-1-methyl-1H-pyrazol-4-yl)-2-(2,6-difluorophenyl)thiazole-4-
carboxamide as the free base (6.3 g, 89%). Purity = 100%; m.p.: 185.73 °C; [α]D20 (deg cm3
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
−
1
−1
−3
1
g dm ) = +6.312 (c = 0.00301 g cm in methanol); H NMR (400 MHz, DMSO-d ) δ 8.84
6
(s, 1H), 7.57 – 7.45 (m, 1H), 7.53 (s, 1H), 7.49 (s, 2H), 7.32 – 7.22 (m, 2H), 4.42 (m, 1H),
3.64 (s, 3H), 3.24 – 3.02 (m, 5H), 2.18 – 2.01 (m, 1H), 2.01 – 1.73 (m, 4H), 1.65 – 1.52 (m,
13
1
2
H); C NMR (101 MHz, DMSO-d ) δ 162.52, 159.86 (dd, J = 6.2, 251.9 Hz), 158.68 (t, J =
6
.6 Hz), 142.04, 133.91 (t, J = 4.0 Hz), 132.56, 131.56 (t, J = 10.8 Hz), 122.71, 113.10 –
112.85 (m), 112.67, 111.36 (t, J = 15.9 Hz), 98.39 (d, J = 167.4 Hz), 55.52 (d, J = 20.4 Hz),
19
4
8.86, 46.41 (d, J = 11.2 Hz), 35.93, 33.58 (d, J = 21.8 Hz), 32.73 (d, J = 7.6 Hz); F NMR
+
(
376 MHz, DMSO-d ) δ -111.24, -171.78; HRMS (m/z): [M+1] calcd for C H F N OS,
6
20 23
3
7
4
66.1559; found, 466.1626; elemental analysis (% calcd, % found for C H F N OS): C
20 22 3 7
(51.60, 51.23), H (4.76, 4.5), F (12.24, 11.89), N (21.06, 20.81), O (3.44, oxygen could not be
analyzed in the presence of fluorine) , S (6.89, 6.67).
But-3-en-1-amine (27). 4-bromobut-1-ene (660 g, 4.9 mol) was added to 18.0 L of 12 M
ammonia in MeOH. Then the mixture was stirred at 90 C in an autoclave overnight. The
mixture was cooled to room temperature and the solvent was removed by concentration. The
crude product was obtained as white solid (660 g, yield 88%).
ACS Paragon Plus Environment