Discovery of novel triazolo[4,3-b]pyridazin-3-yl-quinoline derivatives as PIM inhibitors

Article abstract of DOI:10.1016/j.ejmech.2019.02.022

PIM kinase family (PIM-1, PIM-2 and PIM-3) is an appealing target for the discovery and development of selective inhibitors, useful in various disease conditions in which these proteins are highly expressed, such as cancer. The significant effort put, in the recent years, towards the development of small molecules exhibiting inhibitory activity against this protein family has ended up with several molecules entering clinical trials. As part of our ongoing exploration for potential drug candidates that exhibit affinity towards this protein family, we have generated a novel chemical series of triazolo[4,3-b]pyridazine based tricycles by applying a scaffold hopping strategy over our previously reported potent pan-PIM inhibitor ETP-47453 (compound 2). The structure-activity relationship studies presented herein demonstrate a rather selective PIM-1/PIM-3 biochemical profile for this novel series of tricycles, although pan-PIM and PIM-1 inhibitors have also been identified. Selected examples show significant inhibition of the phosphorylation of BAD protein in a cell-based assay. Moreover, optimized and highly selective compounds, such as 42, did not show significant hERG inhibition at 20 μM concentration, and proved its antiproliferative activity and utility in combination with particular antitumoral agents in several tumor cell lines.

Full text of DOI:10.1016/j.ejmech.2019.02.022

Accepted Manuscript
Discovery of novel triazolo[4,3-b]pyridazin-3-yl-quinoline derivatives as PIM inhibitors
Sonia Martínez-González, Sonsoles Rodríguez-Arístegui, Cristina Ana Gómez de la
Oliva, Ana Isabel Hernández, Esther González Cantalapiedra, Carmen Varela, Ana
Belén García, Obdulia Rabal, Julen Oyarzabal, James R. Bischoff, Javier Klett, María
Isabel Albarrán, Antonio Cebriá, Nuria Ajenjo, Beatriz García-Serelde, Elena Gómez-
Casero, Manuel Cuadrado-Urbano, David Cebrián, Carmen Blanco-Aparicio, Joaquín
Pastor
PII:
S0223-5234(19)30136-9
DOI:
https://doi.org/10.1016/j.ejmech.2019.02.022
EJMECH 11114
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 14 November 2018
Revised Date: 24 January 2019
Accepted Date: 7 February 2019
Please cite this article as: S. Martínez-González, S. Rodríguez-Arístegui, Cristina.Ana.Gó. de la Oliva,
A.I. Hernández, Esther.Gonzá. Cantalapiedra, C. Varela, Ana.Belé. García, O. Rabal, J. Oyarzabal,
J.R. Bischoff, J. Klett, Marí.Isabel. Albarrán, A. Cebriá, N. Ajenjo, B. García-Serelde, E. Gómez-
Casero, M. Cuadrado-Urbano, D. Cebrián, C. Blanco-Aparicio, Joaquí. Pastor, Discovery of novel
triazolo[4,3-b]pyridazin-3-yl-quinoline derivatives as PIM inhibitors, European Journal of Medicinal
Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.02.022.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Discovery of Novel Triazolo[4,3-b]pyridazin-3-yl-quinoline Derivatives as
PIM Inhibitors
a
a
Sonia Martínez-González , Sonsoles Rodríguez-Arístegui , Cristina Ana Gómez de la
a
a
a
a
Oliva , Ana Isabel Hernández , Esther González Cantalapiedra , Carmen Varela , Ana
a
a,1
a,1
a,2
a
Belén García , Obdulia Rabal , Julen Oyarzabal , James R. Bischoff , Javier Klett ,
a
a
a,3
a,4
María Isabel Albarrán , Antonio Cebriá , Nuria Ajenjo , Beatriz García-Serelde , Elena
a
a
a,5
a
Gómez-Casero , Manuel Cuadrado-Urbano , David Cebrián , Carmen Blanco-Aparicio ,
a,*
Joaquín Pastor
a
Experimental Therapeutics Programme, Spanish National Cancer Research Centre
(CNIO), C/ Melchor Fernández Almagro 3, E-28029 Madrid, Spain
Corresponding Author
*
Joaquín Pastor. Phone: +34 917328000, fax: +34 917328051. E-mail address:
jpastor@cnio.es.
Present addresses
1
Small Molecule Discovery Platform. Molecular Therapeutics Program, Center for Applied
Medical Research, CIMA, University of Navarra, Avda. Pío XII, 55; 31008 Pamplona,
Spain.
2
Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd, Grenzacherstrasse 124, 4070
Basel, Switzerland.
3
Tumour Bank Unit, Spanish National Cancer Research Centre (CNIO), C/Melchor
Fernández Almagro 3, E-28029 Madrid, Spain.

ACCEPTED MANUSCRIPT
4
Life Length. Parque Científico de Madrid. C/ Faraday, 7. Campus de Cantoblanco. E-
8049 Madrid, Spain.
2
5
Pharmacology unit, Diseases of the developing world, GSK, Science to Business
Technology Park, C/ Severo Ochoa, 2, E-28760, Madrid, Spain.
Abbreviations
The term AML means Acute myeloid leukemia, CRC means colorectal cancer, NSCLC
means non small cell lung carcinoma, MCL means Mantle cell lymphoma.

ACCEPTED MANUSCRIPT
Abstract
PIM kinase family (PIM-1, PIM-2 and PIM-3) is an appealing target for the discovery and
development of selective inhibitors, useful in various disease conditions in which these
proteins are highly expressed, such as cancer. The significant effort put, in the recent years,
towards the development of small molecules exhibiting inhibitory activity against this
protein family has ended up with several molecules entering clinical trials. As part of our
ongoing exploration for potential drug candidates that exhibit affinity towards this protein
family, we have generated a novel chemical series of triazolo[4,3-b]pyridazine based
tricycles by applying a scaffold hopping strategy over our previously reported potent pan-
PIM inhibitor ETP-47453 (compound 2). The structure-activity relationship studies
presented herein demonstrate a rather selective PIM-1/PIM-3 biochemical profile for this
novel series of tricycles, although pan-PIM and PIM-1 inhibitors have also been identified.
Selected examples show significant inhibition of the phosphorylation of BAD protein in a
cell-based assay. Moreover, optimized and highly selective compounds, such as 42, did not
show significant hERG inhibition at 20 µM concentration, and proved its antiproliferative
activity and utility in combination with particular antitumoral agents in several tumor cell
lines.
Keywords: anticancer agents; selective PIM-1/PIM-3 inhibitors; pan-PIM inhibitors; PIM-1
inhibitors; chemical probes; antiproliferative activity; synergistic effects.
Highlights:
•
Novel potent and highly selective tricyclic PIM inhibitors.

ACCEPTED MANUSCRIPT
•
•
•
•
Mainly PIM-1/PIM-3 profile; pan-PIM and PIM-1 inhibitors also identified.
Chemical probes to determine PIM isoforms contribution in cancer cell viability.
Synergism of inhibitor 42 in combination with several anticancer agents.
Acceptable in vivo Clearance of selected inhibitors to be used in efficacy studies.

ACCEPTED MANUSCRIPT
1
. Introduction
Important roles in cell proliferation, inhibition of apoptosis, cell cycle regulation,
survival, and migration are played by the Proviral Insertion site in Moloney murine
1
-3
leukemia virus (PIM) protein family. Three serine/threonine kinase isoforms comprise
the PIM kinase family: PIM-1, PIM-2 and PIM-3 that are regulated through transcription
4
,5
and stability by JAK/STAT pathways. They have similar functions and have a highly
6
homologous kinase domain, sharing 61-76% of amino acid identity, including the unique
7
feature of having a proline in the hinge region. However, they differ in terms of their
distribution in tissues and organs. PIM-1 and PIM-2 are mainly expressed in hematopoietic
8
cells, while PIM-3 is more abundant in the brain, kidneys, and the epithelia.
Overexpression or dysfunction of PIM kinases has been observed in many types of
hematological malignancies and solid tumors. The three family members have different
levels of expression in various tumor types. PIM-1 is overexpressed in both B-cell non-
Hodgkin’s lymphoma and in myeloid leukemia (AML), in addition to other solid tumors
such as prostate cancer. PIM-2 is mostly expressed in leukemia and lymphomas, while
PIM-3 is implicated in non-hematologic cancers like melanoma, pancreatic and gastric
8
-11
tumors, such as colon or stomach.
Furthermore, different studies have linked the
overexpression of PIM kinases with poor prognosis in gastric cancer and squamous cell
1
2-14
carcinoma of the head and neck.
In vitro and in vivo studies have confirmed PIM
kinases as weak oncogenes individually, but very potent in cooperation with C-MYC in
1
5
certain tumor types. In female hormonal tumor types as breast and uterus PIM-1 and PIM-
1
6
2
behave as oncogene by their own. Silencing PIM-1 induces growth inhibition of
prostate cancer cells,17 hepatocellular carcinoma cells, salivary gland adenocystic
18
carcinoma cells, and mesothelioma cells.20 In addition
, shutting the PIM-3 expression can
1
9

ACCEPTED MANUSCRIPT
2
1
22,23
24
retard in vitro cell proliferation of hepatocellular, pancreatic,
colon carcinoma, and
2
5
gliobastoma cell lines by promoting cell apoptosis. On the other hand, triple PIM knock-
1
5, 26-29
out mice are viable, therefore suggesting minimal side effects for pan-PIM inhibition.
These precedents have triggered the development of PIM inhibitors. The first generation
3
0,31
of this class of drug candidates was focused on imidazo[1,2-b]pyridazine derivatives,
3
2
represented by SGI-1776 (Fig. 1), a pan-PIM inhibitor with nanomolar activity that
exhibited additional inhibition against FLT3 kinase. This molecule was halted in early
clinical evaluation due to a narrow therapeutic window against cardiotoxic side effects
linked to hERG inhibition. Other small molecules identified later, such as the pan-PIM
3
3
34
inhibitors PIM447 (LGH447, Novartis) and INCB053914 (Incyte), are currently being
studied in early clinical trials. It should be mentioned that several of these inhibitors have
shown a greater antitumor effect when combined with other therapies than as individual
3
5-38
agents.
These results point to PIM family members as promising anti-cancer targets for the
development of drugs with favorable on-target toxicity profiles.
Applying a scaffold hopping strategy based on the use of a chemically feasible fragments
database, and the reported structure of SGI-1776, our group developed novel chemotypes
3
9-42
for successful inhibition of PIM kinases.
First we explored a series of triazolo[4,5-
b]pyridines, identifying the early lead compound 2 (ETP-47453), as a pan-PIM inhibitor
Fig. 1a). The X-ray co-crystal structure of a representative inhibitor of this class, ETP-
6546 (1, Fig. 1a) in the catalytic site of PIM-1 unveiled its binding mode and key
4
0
(
4
interactions with the protein (Fig. 1b, PDB: 4A7C). The nitrogen of the triazole ring
establishes a hydrogen bond with Lys67. The 3-trifluoromethoxyphenyl ring makes contact
with the hinge residue Glu121 producing weak H-bond interactions between its C-5 and C-

ACCEPTED MANUSCRIPT
6
protons and the backbone carbonyl group of Glu121. The m-OCF
3
group creates an
Finally, the basic
4
3,44
additional hydrophobic interaction with the hinge residue Pro123.
amine moiety is oriented towards the solvent area participating in an electrostatic
interaction with Asp128.
a)
O
1
7
N
N
N
6
5
N
N
N
N
N
N
N
N
N
N
N
H
H
N
HN
H
HN
CF O
CF O
3
3
CF O
3
SGI-1776
1 (ETP-46546)
2 (ETP-47453)
PIM-1 IC50: 7 nM
PIM-1 IC50: 6 nM
PIM-1 IC50: 1 nM
PIM-2 IC50: 363 nM
PIM-3 IC50: 69 nM
PIM-2 IC50: 230 nM
PIM-3 IC50: 7 nM
PIM-2 IC50: 29 nM
PIM-3 IC50: 0.8 nM
b)
Fig. 1. (a) Chemical structure of PIM inhibitors SGI-1776, 1 (ETP-47546) and 2 (ETP-
7453). (b) Reported crystal structure of PIM-1 with compound 1 (PDB 4A7C). Major
4
interactions marked in yellow.
Now, we present a novel series of tricyclic triazolopyridazine PIM-1 inhibitors (3) (Fig.
2
), which was designed based on experimental PIM-1 activity data of chemically related

ACCEPTED MANUSCRIPT
compounds. First, the PIM-1 activity of our C7-OMe and C6-Cl substituted inhibitors (2,
and compound 9c from reference 40). Next, the possibility of exchanging the triazolo[4,5-
b]pyridine scaffold (compounds
triazolo[4,3-b]pyridazine (compound 27 from reference 44) or triazolo[4,3-a]pyridine
example 9 from reference 46), while keeping the PIM-1 activity (Fig. 2). The activity of
example 9” is of special interest since it represents a different substitution pattern with
1−2) by imidazo[1,2-b]pyridazine (SGI-1776),
(
“
respect to the other PIM-inhibitors.
We performed computational modeling studies with these inhibitors in order to
understand their binding modes and the key interactions with PIM-1. The disubstituted
4
4
compounds SGI-1776, and compound 27 , as well as the monosubstituted inhibitor
4
6
“
example 9 ” (Fig. 2) were docked directly in the X-ray structure 4A7C (Fig. S3.1). On the
other hand, we performed molecular dynamic studies with trisubstituted compounds (2 and
4
0
9
c
from Fig. 2), since the direct docking pointed to potential steric clash of these
compounds with the P-loop residue Phe49 (Fig S3.2). It is known that this part of PIM-1 is
45
highly flexible, therefore it is possible that these inhibitors could be accommodated in the
catalytic site of the protein by displacement of the P-loop. The results of the computational
studies showed that all the inhibitors share the same binding mode as compound 1. In the
case of the trisubstitued inhibitors a clear movement of the P-loop is appreciated.
The visual inspection of the modeled structures, allowed to easily distinguish the
fragments in the molecules which are placed in the Asp128 / Solvent Area (cyclic amine
moieties in blue), the hinge interacting fragments (green), the scaffold N targeting the
Lys67 residue (orange), and the functionality pointing towards the P-loop (highlighted in
red). Two main conclusions emerged from this analysis. First, the model of “example 9”

ACCEPTED MANUSCRIPT
confirmed that the Asp128 / Solvent Area can be accessed from the hinge fragment of such
inhibitor. Second, the induced fit models of C7-OMe and C6-Cl substituted inhibitors
proved that P-loop directed substitutions are allowed, and inspired a potential linking of
such positions by forming a fused morpholinyl cycle (C7-C8 in the tricyclic scaffold 3).
The combination of both features led to the final design of the tricyclic class of PIM
inhibitors presented herein (3, Fig. 2). Interestingly, this tricyclic scaffold represented a
novel unexplored chemical space.
H
1
7
H
N
6
N
5
N
N
N
HN
N
4
N
N
N
HN
N
HN
N
HN
CF O
3
N
N
H
CF O
F3C
3
1/ PDB code: 4A7C
PIM-1, IC50 = 6.0 nM
SGI-1776/ PIM-1, IC50 = 7.0 nM
Compound 27 in ref. 44 / PIM-1 , Ki = 21 nM
O
1
8
N
N
R
7
6
N
N
H
N
4
3
OMe
Hinge
Fragment
N
Asp-128
Solvent
Fragment
N
N
N
N
N
HN
N
N
H N
2
N
HN
Cl
N
CF O
3
N
N
HN
N
2/ PIM-1, IC50 = 1.0 nM
Example 9 in ref. 46 / PIM-1, IC50 = 11 nM
HN
CF O
3
Compound 9c in ref. 40 / PIM-1 , Ki = 6.0 nM
From PDB code 4A7C and Modeling of represented molecules:P-loop oriented fragment Asp-128-Sovent Area Hinge Area Lys-67
Fig. 2. Chemical structures of selected PIM-1 inhibitors. Assignment of structural motifs to
areas / interactions within the PIM-1 catalytic site.
2
. Results and discussion

ACCEPTED MANUSCRIPT
2
.1. Chemistry
The initial exploration of our tricyclic scaffold was focused on finding the best
linker/hinge fragment at its 3-position to attach the Asp128-Solvent targeted moiety.
Several replacements of the 3-trifluoromethoxyphenyl fragment of SGI-1776 and
4
0
39
compound 2 were investigated, from aryl, heteroaryl and arylcarboxamides identified
4
6
previously by us, up to the quinoline moiety of “example 9”.
The preparation of the proposed tricycles was undertaken following the synthetic route
summarized in Scheme 1. The different types of derivatives prepared, had the bicycle 5a
(8-chloro-4-methyl-3,4-dihydro-2H-pyridazino[4,5-b][1,4]oxazine)
as
common
intermediate. Itwas synthesized from commercial trichloropyridazine 4 in two steps. Thus,
nucleophilic substitution at room temperature with 2-methylamino ethanol in methanol,
allowed the selective introduction of the amine at 4-position of the pyridazine ring. The
subsequent cyclization reaction was carried out in the presence of potassium tert-butoxide
in tetrahydrofurane, under high dilution conditions in order to favor the intramolecular
reaction, affording the bicyclic compound 5a. A first approach to prepare
triazolopyridazine tricycles involved the cyclisation of 5a with formic hydrazide, and the
consecutive brominating step to give compound 8, which allowed the introduction of a
suitable aromatic hinge fragment through a final Suzuki cross-coupling. This pathway
enabled the preparation of the derivative 2-fluoro-4-pyridine intermediate 9 that by
nucleophilic aromatic substitution with 4-amino-boc-piperidine in butanol, and further Boc
group deprotection under acidic conditions, gave compound 17.
An alternative synthetic approach was based on the use of intermediate hydrazine bicycle
6
a obtained from 5a by reaction with hydrazine hydrate. The reaction of 6a with
commercial available, or synthesized phenyl, pyridyl, and quinoline carboxaldehydes,

ACCEPTED MANUSCRIPT
afforded tricycles 10−15a with the different hinge fragments and functional groups (Z) for
further derivatization. The versatility of Z groups allowed the introduction of diverse
solvent moieties using different chemical reactions. Thus, the direct transformation of the
carboxylic acid methyl ester group of intermediates 10 and 11 (Z = CO Me) by the use of
2
trimethylaluminum, allowed the introduction of the amines present in tricycles 18 and 19
(Table 1). Additionally, compounds where Z=Br enable the use of palladium-catalyzed
coupling reactions. Compound 12 went under Buchwald-Hartwig coupling conditions
giving access to compound 20, meanwhile reaction of compound 13 with 3-
aminophenylboronic under Suzuki-Miyaura reaction afforded compound 13b which
subsequently reacted with N-Boc-4-piperidine acetaldehyde and N-Boc-4-piperidone by
reductive amination to generate derivatives 21 and 22. Similar reductive amination
conditions were used with tricycle compound 16a (Z = NH ) to obtain compound 27, while
2
conventional amino-acid coupling reaction of 16a with adequate carboxylic acid resulted in
product 28 (Table 1). Given the feasibility to introduce chemical diversity in the solvent
area via reductive amination, more analogues were synthesized using this strategy (29-31,
33−35 and 37−38; Table 2). On the other hand, intermediate 14a with Z= Br allowed the
introduction via Buchwald-Hartwig palladium-catalyzed coupling of a diverse set of
primary and secondary amines, difficult to prepare by other routes, that led to compounds
2
6, 32, 36 (Scheme 1, Tables 1 and 2). The use of acidic media with the ion exchange resin
Amberlyst®15 resin, allowed the Boc group removal for the synthesis of desired analogues
under soft conditions.
Moreover, reaction of 6a with diethyloxalate enabled the preparation of the carboxylic
acid ethyl ester intermediate 7, which was directly transformed into the carboxamide

ACCEPTED MANUSCRIPT
derivatives 23 to 25 (Scheme 1, Table 1) by reaction with selected anilines using the Lewis
acid, trimethylaluminum, and avoiding the use of an alternative two-step transformation
4
7
(carboxylic ester hydrolysis and classical amine coupling).
The substitution of the N atom of the fused morpholinyl ring was additionally explored in
2
the quinoline subseries with alternative alkyl substitutions (R
≠
Me). The bicycles 5b-f
2
were prepared from trichloropyridazyine 4 by reaction with the substituted R -ethanol
amines, and subsequent intramolecular cyclization. The next synthetic transformations to
obtain the final tricycles (39−43; Scheme 1, Table 2) followed similar chemical conditions
2
as described for compound 27 (R = Me).
O
O
O
N
N
N
N
N
N
e
g, h, j, or k and m
Me
Me
Me
N
N
N
N
N
N
N
N
N
Br
X
X
8
9-13b
17-22
Y
1
Y
z
R
9, X = C; Y = N; Z = F
1
0, X = C; Y = C; Z = CO Me
2
1
1, X = N; Y = C; Z = CO Me
2
d
f
12, X = C; Y = C; Z = Br
3, X = N; Y = C; Z = Br
1
i
13b, X = N; Y = C; Z =
NH2
Cl
Cl
NH2
HN
O
O
N
Cl
Cl
O
N
O
N
N
N
a
N
N
b
c
h
N
N
N
N
N
N
Me
Me
N
N
N
N
2
N
2
R
R
O
O
4
5a-f
6a-f
7
EtO
23-25 HN
1
a, R2 = Methyl
R
b, R2 = Ethyl
c, R2
d, R2
e, R2
f, R2
=
=
=
f
OPh
OMe
=
O
N
z
O
2
N
N
Z = NH
2
, j, and m
2
N
N
R
R
N
N
N
N
Z = Br, k or l, and m
N
N
N
1
R
14a, Z = Br
1
5a-f, Z = NH-Boc
R2
R
=
≠
Methyl, 26-38
Methyl, 39-43
m
2
16a-f, Z = NH2

ACCEPTED MANUSCRIPT
Scheme 1. Synthesis of triazolo[4,3-b]pyridazines tricycles analogues. Reagents and
t
conditions: (a) i) R
8-87%; (b) NH NH
ºC, 3 h, 52%; (d) i) formic hydrazide, nBuOH, 150 ºC, 24 h, 65-71%, ii) NBS, CHCl , rt,
2
NHCH
2
CH
2
OH, MeOH, rt, 1h, 27-57%, ii) BuOK, THF, 65 ºC, 1-2 h,
1
2
2
.H
2
O, 150 ºC, microwave, 45 min, 62-94%; (c) diethyl oxalate, 130
3
3
6 h, 25%; (e) 2-fluoropyridine-4-boronic acid, PdCl (dppf), K CO (aq), DME, 130 ºC,
2 2 3
microwave, 60 min, 51% (f) i) Aryl/heteroaryl aldehyde, EtOH, 76 ºC, 1 h, ii)
diacetoxyiodo)benzene, DCM, rt, 1-6 h, 46-87% over two steps; (g) i) 4-amino-Boc-
piperidine, nBuOH, 200 ºC, 24-48 h, 30-40%; ii) HCl, dioxane, rt, 18 h, 28%; (h) Amine,
AlMe , THF, DCM, 40 ºC, 18 h, 36-56%; (i) 3-aminophenylboronic acid, PdCl (dppf),
CO (aq), DME, 130 ºC, microwave, 30 min, 68%; (j) Aldehyde or ketone, AcOH, DCE,
NaBH(OAc) , rt, 18 h, 10-74%; (k) Amine, Pd (dba) , DavePhos, NaOtBu, dioxane, 130
ºC, 40-90 min, microwave, 10-25%; (l) Amine, Pd (dba) , rac-BINAP, Cs CO , toluene,
(
3
2
K
2
3
3
2
3
2
3
2
3
1
1
20 ºC, 12-48 h, 5-11%; (m) Amberlyst®15, MeOH, rt, 24 h, 30-90%. For definitions of R
2
and R see Table 1 and 2.
3
The preparation of derivatives with additional small groups substitutions (R = F, Cl, Me)
in the quinoline ring is depicted in Scheme 2. It was carried out by cyclization of hydrazine
3
48
bicycle 6a, and the readily prepared 8-bromo-R -quinoline-2-carboxaldehydes 44a-f, to
afford tricycles intermediates 45a-f. The subsequent reaction with tert-butyl 4-
(aminomethyl)-4-fluoropiperidine-l-carboxylate under Buchwald coupling conditions
proceeded in moderate yields to give analogues 46 51 after Boc-group removal.
−
The replacement of 2-quinoline moiety by isoquinoline and naphthalene was achieved by
condensation of the common hydrazine intermediate 6a with the chloro substituted bicyclic
aldehydes 52 and 53, to obtain the corresponding tricycles, 54 and 55. The different

ACCEPTED MANUSCRIPT
“solvent targeting groups” were subsequently introduced using several amines by
nucleophilic aromatic substitution to yield the isoquinoline derivatives, 56
Buchwald palladium catalyzed coupling to afford the naphthalene analogues 59
Scheme 2, Table 2).
−
58, and by
−61
(
Br
O
O
N
O
N
N
7
H
N
N
N
N
Me
Me
6
3
R
N
N
N
N
5
a
b
4
4a-f
a, R3 = 6-F
N
N
3
H
b, R = 7-Cl
c, R3 = 6,5-F
Br
HN
N
5
5
3
3
3
d, R = 6-Cl
7
R
F
7
R
3
6
6
e, R = 5-F
3
f, R = 6-Methyl
45a-f
46-51
NH2
HN
O
N
N
N
Me
6a
Cl
O
Y
H
O
O
N
N
N
N
Me
N
5
2, Y = N
Me
N
N
a
53, Y = C
N
c
N
N
1
R
Cl
Y
Y
5
5
4, Y = N
5, Y =C
56-58, Y = N
59-61, Y = C
Scheme 2. Synthesis of additional triazolo[4,3-b]pyridazines tricycles analogues. Reagents
and conditions: (a) i) EtOH, 76 ºC, 1-2 h; ii) (diacetoxyiodo)benzene, DCM, rt, 1-3 h, 28-
5
0% over the two steps (b) i) tert-butyl 4-(aminomethyl)-4-fluoropiperidine-l-carboxylate,
Pd (dba) , rac-BINAP, Cs CO , dioxane, 110 ºC, 20-24 h; ii) Amberlyst, MeOH, rt, 24 h,
2-48% (c) for Y = N, i) Amines, TEA, nBuOH, 180 ºC, 18-24 h; ii) only for Boc
deprotection: Amberlist, 28-45% overall yield; for Y = C, i) Amines, Pd (dba) , Cs CO ,
2
3
2
3
1
2
3
2
3
Brettphos or X-phos, dioxane, 110 ºC; ii) only for Boc deprotection: HCl 4N/dioxane, rt.
1
3
2
6-49% overall yield. For definitions of R and R see Table 2.

ACCEPTED MANUSCRIPT
2
2
.2. Biological evaluation
.2.1. PIM-1 inhibition and Structure Activity relationship (SAR) studies
In order to test our hypothesis we evaluated in dose response assays the inhibitory activity
(
IC50 value, half maximal inhibitory concentration) of the prepared compounds against
PIM-1 (Table 1).
The examination of the PIM-1 inhibition results of initial tricyclic derivatives 17−28
revealed that the quinoline hinge fragment clearly yielded the most potent compounds, 27
IC50 = 14 nM) and 28 (IC50 = 36 nM) with low nanomolar inhibitory activities comparable
(
to the reference bicycles SGI-1776 and 2. The analogue 26, with a piperazine moiety, a two
atoms shorter “Asp128-Solvent targeted residue”, showed moderate PIM-1 inhibition, IC50
=
200 nM, 14-fold less potent than 27 and 28. In contrast to quinoline derivatives, the
compounds 17 21 with a pyridyl and phenyl hinge fragment, displayed significantly
weaker activity (IC50 approximately 1µM), and only the derivative 22 rendered an
interesting submicromolar activity (IC = 425 nM), but 30-fold less potent than compounds
−
≥
5
0
2
7 and 28. Unfortunately, the products bearing arylcarboxamides hinge fragments (23−25),
potential bioisosters of the quinoline moiety, were much less active. Only compound 25
1
with R = 4-oximethylpiperidine that could establish an internal H-bond between the
1
oxygen atom of R and the NH moiety of the carboxamide group, was able to exhibit some
modest PIM-1 inhibitory activity (IC50 = 1.1 µM) in comparison with its close inactive
analogue the 4-amine-methyl-piperidine 23 (IC > 10 µM).
5
0
As a result of this initial exploration, we identified tricycles 27−28 as low nanomolar
PIM-1 inhibitors. Computational modeling of the binding mode of inhibitor 27 in PIM-1,
through the use of molecular dynamics, confirmed that proposed in our design (Fig.3). The

ACCEPTED MANUSCRIPT
N of the triazole moiety of the scaffold established an H bond with Lys67. The quinoline is
accommodated in the hinge area. The 4-aminomethyl piperidine basic fragment pointed to
the solvent accessible region where it produced an electrostatic interaction with Asp128.
Finally, the fused morpholine cycle, whose O atom established an additional H bond with
Lys67, is placed under the P-loop which is displaced in comparison with structure 4A7C
for the bicyclic inhibitor 1 (Figure 3A, B).
Fig. 3. A) Computational model of the binding mode of tricyclic compound 27 in PIM-1.
B) Overlay with structure 4A7C (grey color). Displacement of Phe49 residue of P-loop by
compound 27 is observed (yellow arrow).
Together, experimental activity data and molecular modeling studies confirmed our
design hypothesis. First, that the access to the solvent region growing from C3 of the
designed tricyclic triazolopyridazines was feasible. Second, the accommodation of the
designed tricyclic scaffold in the catalytic site of PIM-1 was possible due to the flexibility
of its P-loop. The SAR of the initial set of tricycles 17−28 can be explained as the result of
the contribution of the different hinge fragments to the proper orientation of the “Asp128-

ACCEPTED MANUSCRIPT
Solvent targeted residue” within the PIM-1 catalytic site, as well as the optimal distance
and architecture of the group bearing a basic interacting moiety in these derivatives.
Next, the quinoline fragment (structure C, Table 1) was selected for further investigation
to establish additional SAR around this novel tricycle scaffold and evaluate its potential to
generate advanced PIM inhibitors.
Table 1
a
PIM-1 inhibitory activity of triazolo[4,3-b]pyridazines.
O
O
O
N
N
N
N
N
N
Me
Me
Me
N
N
N
N
N
N
N
N
N
O
X
HN
N
1
Y
1
1
R
R
R
A
B
C
1
PIM-1
IC50(µM)
1
PIM-1
IC50(µM)
Nr
R
X
-
Y
-
Structure
Nr
R
Structure
SGI-
0.013 b
c
-
-
2
-
-
0.001
>10
1
776
(0.007)
NH
NH
H
1
1
1
2
7
8
9
0
C
N
A
A
>10
0.870
7.1
23
24
25
26
N
B
N
H
O
O
NH
N
H
C
N
C
C
C
C
B
B
C
>10
N
NH
H
NH
N
H
A
A
1.1
O
NH
NH
NH
H
N
>10
N
0.200
NH
H
N
NH
NH
H
N
2
1
N
N
C
C
A
A
0.785
0.425
27
28
C
C
0.014
0.036
H
N
H
N
2
2
NH
O
a
b
The values reported are an average of two independent data points. In brackets data for
c
SGI-1776 in accordance with ref. 32. Data in accordance with ref. 40.
1
First, the effect of the different “solvent targeting residues” (R ) in the triazolo[4,3-
b]pyridazin-3-yl-quinoline tricycles was evaluated. Compound 26 with a piperazine moiety,
6
-membered ring, was significantly less active (PIM-1 IC = 200 nM) than compounds
50
with longer and flexible solvent moieties 27 (PIM-1 IC = 14 nM) and 28 (PIM-1 IC =
5
0
50

ACCEPTED MANUSCRIPT
3
6 nM), then we synthesized regioisomers 29 and 36 with 4-aminopiperidine a fragment of
intermediate size. Both compounds behaved as good PIM-1 inhibitors (29: IC50 = 12 nM
and 36: IC50 = 38 nM, Table 2). Next, inhibitors 27 and 29 were methylated at their N-
terminal atoms affording compounds 31 and 30 which maintained or even improved the
PIM-1 activity (IC50 = 14 and 7 nM, respectively). Therefore, it seems that the methyl
group is perfectly accommodated in this area of PIM-1 still allowing the electrostatic
interaction with Asp-128 or nearby residues. Next, we decided to explore the importance of
a basic moiety in the solvent residue. Aminopiperidine from compound 29 was replaced by
aminotetrahydropyranyl (33) and aminocyclohexyl (35) fragments. Both compounds were
good PIM-1 inhibitors (IC50 = 25 and 32 nM) observing only a slight decrease in activity of
2
-3 folds comparing to compound 29. A similar trend was observed in the replacement of
the aminomethylpiperidine residue of compound 27 (IC50 14 nM) by
=
aminomethyltetrahydropyranyl (34; IC50 = 30 nM). However, the replacement with
aminomethylcyclohexyl (38) led to a significant drop in PIM-1 activity (IC = 420 nM). A
5
0
slight decrease in PIM-1 activity was also observed in compound 37, when a gem-dimethyl
moiety was placed in the cyclohexyl fragment (IC = 92 nM). These results could indicate
5
0
that “longer/bigger” lipophilic aliphatic fragments produce unfavorable interactions in this
highly hydrophilic area of the protein, resulting detrimental for the activity of the
compounds 37 and 38. Interestingly, the fine tuning of basicity in 32 by introduction of a
fluorine atom at the 4-position of piperidine, did not represent a major impact in their
activity (IC = 21 nM) in comparison with 27 (IC = 14 nM). This finding was of special
5
0
50
4
0
interest since compound 2 and SGI-1776 showed undesirable hERG binding activity and
4
9
as it is reported in literature the attenuation of basicity (i.e. 4-F Piperidine substitution)
could contribute to avoid it in our tricyclic inhibitors. In addition, the introduction of other

ACCEPTED MANUSCRIPT
2
2
alkyl residues (R ≠ Me) to study the influence of the R substituent at the morpholinyl
fused ring (compounds 39−43, Table 2) revealed that this position of the tricycle seemed to
be tolerant even to bulky fragments such as phenoxyethyl in compound 39 (IC50 = 17 nM)
and a branched substituent such as isopropyl (43, IC50 = 16 nM). These results seem to
reinforce the fact that the third ring of these tricyclic inhibitors is well accommodated in the
2
P-loop region of PIM-1 where the R substituent could protrude to an open solvent area in
this region.
3
The influence of additional substitutions (R ) in compound 32, which was taken as
reference for this focused SAR study was also investigated. Thus, we prepared compounds
4
6−51 (Table 2). The substitution at the 5-, 6- or 7-positions with halogens (F, Cl), did not
affect the PIM-1 inhibitory activity (compounds 46−50), however 6-methylation (51)
resulted in 10-fold potency decrease (IC50 = 236 nM). This effect could be due to an
unfavorable steric clash of the methyl group with the close Val126 residue in the hinge area
of PIM-1. However, the comparison of 51 with its 6-Cl analogue 49 (IC50 = 30 nM) did not
support such hypothesis since the Van der Waals radii of Cl and Me are rather similar.
Alternatively, the 6-Me substitution could induce an unfavorable conformational effect
disturbing the adequate orientation of the 4-F piperidine moiety whereas the 6-Cl group not.
In addition, other contributions of the 6-Cl group in 49 such as inductive effects or halogen
type interactions could play a role to its activity.
Table 2
SAR exploration of Triazolo[4,3-b]pyridazin-3-yl-quinoline tricycles.

ACCEPTED MANUSCRIPT
O
2
R
N
N
N
N
N
N
1
8
7
R
5
3
6
R
1
2
3
PIM-1 a
PIM-2 a
PIM-3 a
P-BAD b
EC50(nM)
Ratio P-
BAD/PIM-1
Nr
R
R
R
IC50(nM)
IC50(nM)
IC50(nM)
SGI-
13.4
(7)
618 c
50.5c
c
-
-
-
-
-
-
c
200
70
29
70
27
1
776
(362)
(69)
d
2
1
29
0.8
NH
2
2
6
7
N
Me
Me
H
H
200
>10000
257
5360
NH
NH
H
N
14
36
4260
3740
17
700
330
50
9
H
N
2
2
8
9
Me
Me
H
H
131
O
NH
12
690
95
1380
115
N
H
Me
N
3
3
3
0
1
2
Me
Me
Me
H
H
H
7
2870
4250
5530
40
82
55
212
179
215
30
13
10
N
H
Me
N
H
N
14
21
NH
H
N
F
O
3
3
3
3
3
4
5
6
Me
Me
Me
Me
H
H
H
H
25
30
32
38
2470
6210
3330
2140
358
402
300
80
322
325
13
11
7
N
H
O
H
N
220
N
H
NH
2
N
5310
140
3
7
Me
Me
H
92
>10000
4080
N/A
N/D
N
H
H
3
3
4
4
4
4
8
9
0
1
2
3
N
H
H
H
H
H
H
420
17
12
13
7
>1000
4690
4010
1930
5530
1650
>10000
140
60
N/D
2020
800
780
262
915
N/D
119
67
NH
NH
NH
NH
NH
H
N
OPh
H
N
H
N
Et
58
60
H
N
OMe
70
37
H
N
16
76
57
NH
NH
NH
NH
NH
NH
H
N
4
4
4
4
5
6
7
8
9
0
1
Me
Me
Me
Me
Me
Me
6-F
10
12
858
696
7
6
465
71
47
6
F
F
F
F
F
F
H
N
7-Cl
H
N
5,6-
di-F
18
3590
1270
3010
>10000
12
1
155
N/A
283
9
H
N
6-Cl
5-F
30
N/D
9
H
N
30
14
N/D
H
N
5
6-Me
236
5960
25
a
b
The values reported are an average of two independent data points. The values are
averages of two independent experiments performed in duplicate. Assay conditions under

ACCEPTED MANUSCRIPT
c
in cell-Elisa format. In brackets data for SGI-1776 in accordance with ref. 32. N/A: Not
active. N/D: Not Determined.
Furthermore, the replacement of quinoline by isoquinoline (compounds 56−58) or
naphthalene (59−61) using some selected fragments to get access into the solvent area, had
some influence in the PIM-1 inhibition. All the compounds retained the activity below 50
nM (Table 3), and kept or improved the potency versus their corresponding quinoline
analogues.
Table 3
SAR exploration of Triazolo[4,3-b]pyridazin-3-yl-quinazoline/naphtyl tricycles.
O
N
N
N
N
N
1
R
Y
1
PIM-1
a
PIM-2 a
PIM-3 a
P-BAD b
Ratio P-
BAD/PIM-1
Nr
R
Y
N
IC50(nM)
IC50(nM)
IC50(nM)
EC50(nM)
NH
NH
H
N
5
5
6
7
4
7
27
6
2510
200
628
29
H
N
F
N
N
122
26
N
5
8
H
26
1100
238
110
4
NH
H
N
5
6
9
0
C
C
6
4870
9940
58
865
144
O
H
N
20
281
N/D
N/D
6
1
C
46
8040
1550
265
6
N
H
a
b
The values reported are an average of two independent data points. The values are
averages of two independent experiments performed in duplicate. Assay conditions under
c
d
in cell-Elisa format. Data in accordance with ref. 32. Data in accordance with ref. 40.
N/D: Not Determined.

ACCEPTED MANUSCRIPT
2
.2.2. PIM isoforms profile
To complete our SAR study the activity against PIM-2 and PIM-3 isoforms was
evaluated (Table 2 and 3). The results, in general, clearly pointed to a PIM-1/PIM-3
selective inhibitory profile for these novel tricycles, showing relatively weak inhibition
towards PIM-2 compared with PIM-1. Most of the compounds have micromolar IC50
values against PIM-2 with the exception of the isoquinoline derivatives 56 and 57 which
behave more as pan-PIM inhibitors. In particular, compound 56 displayed a significant
PIM-2 activity (IC = 27 nM) in comparison with its quinoline counterpart 27 (IC = 4260
5
0
50
nM) highlighting the importance of the N position in the bicyclic moiety for PIM-2
inhibition in this pair of analogues. Another observed trend is that the compounds lacking a
1
terminal basic moiety in the solvent R fragment (compounds 33−35, 37, 38, 58 and 60−61)
are significantly less potent in PIM-3 versus PIM-1, approximately 10 fold, in addition to a
lack of PIM-2 inhibition (IC > 1000 nM). In particular, compound 61 might represent an
5
0
initial hit to develop isoform selective PIM-1 inhibitors. On the other hand, PIM-3 activity
2
decreases as the substituents in R (compounds 39−43) are bigger.
Interestingly, the 6-Cl quinoline derivative 49, demonstrated to be 30-fold more potent
towards PIM-3 than to PIM-1, with negligible PIM-2 activity, thus representing a potential
avenue to obtain PIM-3 selective inhibitors within this chemical series (Table 2).
2
.2.3. Mechanism of action in cells
The cellular activity of the compounds was evaluated by measuring their effects on the
phosphorylation of the PIM effector protein BAD in the non small cell lung cancer H1299
cell line, which overexpress PIM-1 isoform. BAD, a downstream target of the PIM kinases,

ACCEPTED MANUSCRIPT
has been demonstrated to be directly phosphorylated at Ser112 and Ser136 by PIM-1, PIM-
5
0,51
2
1
and PIM-3.
Nevertheless in our cellular setting, H1299 cell line overexpressing PIM-
, the observed inhibition of BAD phosphorylation should be mostly attributable to the
PIM-1 inhibition exerted by the compounds. It was found that some tricyclic PIM inhibitors
efficiently and dose-dependently inhibited the phosphorylation of BAD at Ser112 with
EC values (half maximal response concentration) in the submicromolar range. In general,
50
it was not observed a strong correlation between PIM-1 biochemical inhibition and the
corresponding modulation of BAD phosphorylation, suggesting that other factors such as
permeability of the inhibitors influence their cellular activity. Compounds within the
quinoline / N-methylmorpholine subseries with a significant biochemical PIM-1 inhibition,
IC50 < 50 nM (27−36), displayed cellular activities which do not directly correlate with
their biochemical potency. Here, it is possible to establish a SAR for the solvent exposed
fragment in these compounds using their corresponding ratios P-BAD EC / PIM-1 IC as
5
0
50
a measure of the correlation between biochemical and cellular activity. The greatest
disagreement (ratio > 100) was observed for compounds 36 and 29 with a 4-
aminopiperidine fragment and explicit H-bond donors in their structures. Compound 27
with the insertion of a C atom between the piperidine fragment and the heterocycle linking
N, improved the correlation (ratio = 50) in comparison with compound 29. The same effect
was observed for its methylated analogue 30 and 31 the methylated analogue of 27. The
replacement of the basic NH group of piperidine in 29 by an O atom in the
tetrahydropyranyl analogue 33, showed to be beneficial for its cell activity (ratio = 13). The
same chemical modification improved the cellular activity of 34 versus 27, even with lower
biochemical PIM-1 potency. As it was expected, the introduction of a more lipophilic and

ACCEPTED MANUSCRIPT
non basic cyclohexyl fragment in compound 35 impacted positively in its cellular activity
(ratio = 7). Interestingly, compounds 28 and 32, improved their activity ratios (28 ratio = 9;
3
2 ratio = 10) with respect to 27. This result could be explained by the reduction of the
basicity of piperidine due to the electron withdrawing effects of acetyl and fluorine groups.
The N-substitution of the fused morpholine ring in derivatives 40, 41 and 43 with n-Pr, Et
and i-Pr respectively, did not afford significant changes neither in the biochemical PIM-1
inhibition nor in the cellular activity in comparison with the N-Me substituted inhibitor 27.
However, substitution with bulkier and more lipophilic phenoxyethyl group in 39 afforded
a much higher disagreement between its cellular and biochemical activity (ratio = 119)
while maintaining its PIM-1 inhibition. Interestingly, the N-methoxyethyl substituted
inhibitor 42 was identified as the most potent compound at biochemical (PIM-1) and
cellular level within this subseries, although the disconnection with its PIM-1 inhibition is
still evident (ratio = 37).
The substituted quinoline analogues 48 (5,6-di-F) and 50 (5-F) did not afford major
changes in comparison with the unsubstituted inhibitor 32. Conversely, the 7-Cl analogue
4
7 yielded better cellular inhibition of P-BAD (71 nM), potentially due to a favorable
lipophilic effect of the Cl atom on its cellular permeability. Surprisingly, the 6-Cl
substitution in compound 49 produced the loss of its cell activity. A weaker detrimental
effect on cell activity was also observed for the 6-F analogue 46 (P-BAD EC = 465 nM).
50
The comparison of pairs of analogues of subseries quinoline / isoquinoline rendered
interesting results. The isoquinoline derivatives 56, 57 and 58 showed to be slightly more
active inhibitors of PIM-1/3 than their corresponding quinoline counterparts 27, 32 and 35.
However, compound 56 showed worse cellular activity than inhibitor 27 (2.5 µM vs. 0.7

ACCEPTED MANUSCRIPT
µM), reflecting a negative contribution of the isoquinoline structure when the solvent
exposed fragment is 4-aminomethylpiperidine. Importantly, this effect was attenuated for
the pair isoquinoline 57 and quinoline 32, which are substituted with 4-F-piperidine (P-
BAD EC = 200 nM vs. 215 nM), although the cellular / biochemical ratio was 3 times
50
worse (57 = 29 vs. 32 = 10). Moreover, the negative effect of the isoquinoline structure in
this class of PIM inhibitors was reversed by the introduction of a more lipophilic
cyclohexyl moiety, as it can be observed by comparison of P-BAD EC50 of pair 58 and 35
(110 nM vs. 220 nM). Next, the naphtyl analogue 59 showed higher cellular / biochemical
ratio than the quinoline 27 (ratio = 144 vs. ratio = 50). Furthermore, the introduction of the
cyclohexyl fragment in this naphtyl derivatives (61) recovered greatly the cellular potency
(265 nM) and improved the correlation with its weaker biochemical PIM-1 activity (PIM-1
IC = 46 nM; ratio = 6).
5
0
2
.2.4. FLT3 kinase and hERG binding activity
The PIM inhibitor SGI-1776 is also a strong FLT3 kinase inhibitor which showed
undesirable hERG inhibition activity. Both activities were evaluated in a representative set
of cell active tricyclic PIM inhibitors (P-BAD EC50 300 nM). The novel tricyclic
≤
triazolopyridazine compounds exhibited higher selectivity against FLT3 (> 100 fold), and
better hERG binding profiles than SGI-1776 (Table 4). The presence of the basic N-
Methyl-4 aminomethyl piperidine fragment in SGI-1776 is thought to be the major
structural determinant for its hERG binding activity and its cardiotoxicity. Interestingly, the
novel tricycle 31 with the same basic fragment showed negligible hERG binding activity.
This result suggested a beneficial influence of the tricyclic scaffold, and the particular

ACCEPTED MANUSCRIPT
substitution pattern of 31, despite its basicity, to avoid potential cardiotoxicity issues.
Along this line, basic compounds 30, 31 and 42 or analogues with attenuated basicity 32,
4
8, 50 and 57 and the non-basic derivative 35, showed hERG IC50 values close to or above
0 µM. However, it is not a general trend within this series of tricyclic inhibitors. So, the
2
analogue 47 with attenuated basicity and non-basic inhibitors 58 and 61 rendered the higher
inhibition values (hERG IC50 between 2 − 6 µM) within this series. These results indicate
that other factors beyond basicity, such as lipophilicity or individual substitution patterns
contribute to the observed hERG inhibition.
Table 4
Characterization of selected compounds.
b
Nr
FLT3 IC (µM) hERG IC (µM)
50 50
a
SGI-1776
2
0.021 (0.044)
0.511
2.1
4.35
3
3
3
3
4
4
4
5
5
5
0
1
2
5
2
7
8
0
7
8
>10
>20
>20
19.7
>20
>20
2.1
>10
>10
>10
>10
>10
6.32
>20
>20
>20
4
3.97
>10
>10

ACCEPTED MANUSCRIPT
6
1
>10
6.06
a
b
In brackets data in accordance with ref. 32. Predictor hERG Assay test kits from
Invitrogen. Fluorescence polarization binding assay.
2
.2.5. Additional characterization of compound 42
According to the results presented above, several compounds showed an attractive
profile, and therefore deserved further characterization. In this regard, compound 42, a
potent PIM1/3 inhibitor (7/70 nM) without FLT3 activity and insignificant inhibition of
hERG binding at 20 µM, was selected as representative of the quinoline subseries.
2
.2.5.1. Broad selectivity of compound 42
The selectivity of 42 against other kinases was evaluated. First, it was tested against a
panel of 25 representative kinases at a single concentration of 1 µM showing no significant
5
2
activity (< 25% inhibition). Next, compound 42 was extensively profiled in a high-
throughput competition binding assay against a panel of 468 kinases (KINOMEscan®) at 1
µM. It demonstrated to be a highly selective compound, therefore, a good and useful
chemical probe against PIM1/3 kinases for further biological studies. Inhibitor 42 only
exhibited < 35% of the signal produced by the untreated controls in this assay (a value
widely accepted for a significant binding in this platform) for three kinases (Fig. 4), their
expected targets PIM-1 and PIM-3 and WNK3 (11% signal versus control).

ACCEPTED MANUSCRIPT
Fig. 4. Selectivity profile of compound 42. Image generated using TREEspot™ Software
Tool and reprinted with permission from KINOMEscan®, a division of DiscoveRx
Corporation, © DISCOVERX CORPORATION 2010.
2
.2.5.2. Antiproliferative activity and cell cycle regulation of compound 42
The antiproliferative activity of 42 was assessed in a broad panel of tumor cell lines,
which cover the most relevant tumor types. The residual viability of cells treated with the
compound for 72 hours in comparison with DMSO treated cells was determined. As it is
shown in Table 5, 42 achieved single-digit micromolar GI50 values (concentration which
causes 50% reduction in cell proliferation) for several cell lines. Some of the sensitive cell
lines belong to different tumor types such as colon, pancreas, Ewing’s sarcorma, bladder
cancer, lung cancer and melanoma. In these cell lines PIM-1 and/or PIM-3 are
overexpressed. The melanoma cell line SKMEL-19 showed the greatest sensitivity for our
PIM-1/3 inhibitor 42 (GI50 = 1.48 µM).

ACCEPTED MANUSCRIPT
Table 5
Antiproliferative activity of compound 42.
GI50
GI50
(µM)
17.8
Cell line
HL60
Tumor type
AML
Cell line
U251
Tumor type
a
a
(
µM)
13.6
19.9
6.46
10.2
19.6
21
Glioblastoma
K562
AML
AML
U87MG
A549
Glioblastoma
NSCLC
21.7
16.2
9.7
MV4:11
MDAMB175 Breast cancer
NCIH522
NCIH1650
JEKO-1
SKMEL-19
A375
NSCLC
MCF-7
SKBR-3
T47D
Breast cancer
Breast cancer
Breast cancer
Breast cancer
NSCLC
22.5
7.3
MCL
10.8
20.4
21.2
9.68
5.25
21.2
8.17
8.73
23.2
12.7
7.38
14.9
Melanoma
1.48
21.8
23.5
7.25
9.21
17.4
25.4
22.4
17.1
18.6
8.98
22.1
BT549
Melanoma
MDAMB231 Breast cancer
SKOV3
ASPC-1
BXPC-3
MIAPaCa-2
PANC-1
DU145
Ovary cancer
Pancreatic cancer
Pancreatic cancer
Pancreatic cancer
Pancreatic cancer
Prostate cancer
Prostate cancer
Prostate cancer
Bladder cancer
Bladder cancer
NAMALWA
COLO205
HCT116
SW480
SW620
T84
Burkitt lymphoma
CRC
CRC
CRC
CRC
CRC
PC3
A673
Ewing sarcoma
Ewing sarcoma
Ewing sarcoma
LNCAP
A498
SK-ES-D
TC71
CAKI-1
a
GI50, Growth inhibition 50, micromolar values for inhibition of proliferation.
The effect of inhibitor 42 on the regulation of the cell cycle was evaluated. SKMEL19
cells were incubated with increasing concentration (1−10 µmol/L) of 42 for 24 hours, and
the resulting cell cycle profile was analyzed by flow cytometry. Compound 42 produced a
clear induction of apoptosis at the highest tested concentration of 10 µM (Fig. 5). However,