Rational Design of Novel Highly Potent and Selective Phosphatidylinositol 4-Kinase IIIβ (PI4KB) Inhibitors as Broad-Spectrum Antiviral Agents and Tools for Chemical Biology

Article abstract of DOI:10.1021/acs.jmedchem.6b01465

Phosphatidylinositol 4-kinase IIIβ (PI4KB) is indispensable for the replication of various positive-sense single stranded RNA viruses, which hijack this cellular enzyme to remodel intracellular membranes of infected cells to set up the functional replication machinery. Therefore, the inhibition of this PI4K isoform leads to the arrest of viral replication. Here, we report on the synthesis of novel PI4KB inhibitors, which were rationally designed based on two distinct structural types of inhibitors that bind in the ATP binding side of PI4KB. These “hybrids” not only excel in outstanding inhibitory activity but also show high selectivity to PI4KB compared to other kinases. Thus, these compounds exert selective nanomolar or even subnanomolar activity against PI4KB as well as profound antiviral effect against hepatitis C virus, human rhinovirus, and coxsackievirus B3. Our crystallographic analysis unveiled the exact position of the side chains and explains their extensive contribution to the inhibitory activity.

Full text of DOI:10.1021/acs.jmedchem.6b01465

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Rational design of novel highly potent and selective
phosphatidylinositol 4-kinase III# (PI4KB) inhibitors as broad-
spectrum antiviral agents and tools for chemical biology
Ivana Mejdrová, Dominika Chalupská, Pavla Pla#ková, Christin Müller, Michal
Sala, Martin Klíma, Adriana Baumlová, Hubert Hrebabecky, Eliska Prochazkova,
Milan Dejmek, Dmytro Strunin, Jan Weber, Gary Lee, Marika Matoušová, Helena
Mertlíková-Kaiserová, John Ziebuhr, Gabriel Birkus, Evzen Boura, and Radim Nencka
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01465 • Publication Date (Web): 08 Dec 2016
Downloaded from http://pubs.acs.org on December 11, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 75
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
Rational design of novel highly potent and selective
phosphatidylinositol 4ꢀkinase IIIβ (PI4KB)
inhibitors as broadꢀspectrum antiviral agents and
tools for chemical biology
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a,b‡
a,‡
a
c
a
Ivana Mejdrová , Dominika Chalupská , Pavla Plačková , Christin Müller , Michal Šála ,
a
a
a
a
a
Martin Klíma , Adriana Baumlová , Hubert Hřebabecký , Eliška Procházková , Milan Dejmek ,
a
a
b
a
a
Dmytro Strunin , Jan Weber , Gary Lee , Marika Matoušová , Helena MertlíkováꢀKaiserová ,
c
d
a,*
a,*
John Ziebuhr , Gabriel Birkus , Evzen Boura , Radim Nencka
a
Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic,
v.v.i, Gilead Sciences & IOCB Research Centre, Flemingovo nám. 2, 166 10 Prague 6, Czech
Republic.
b
Department of Chemistry of Natural Compounds, Institute of Chemical Technology Prague,
Technická 5, Prague 166 28, Czech Republic
c
Institute of Medical Virology, Justus Liebig University Giessen, Schubertstrasse 81, Dꢀ35392
Giessen, Germany
d
Gilead Sciences, Inc., 333 Lakeside Drive, Foster City, CA 94404, United States.
ACS Paragon Plus Environment
1

Page 3 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
KEYWORDS
Phosphatidylinositol 4ꢀkinase IIIβ, picornaviruses, hepatitis C virus, selective inhibitor.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ABSTRACT
Phosphatidylinositol 4ꢀkinase IIIβ (PI4KB) is indispensable for the replication of various
positiveꢀsense single stranded RNA viruses, which hijack this cellular enzyme to remodel
intracellular membranes of infected cells to set up the functional replication machinery.
Therefore, the inhibition of this PI4K isoform leads to the arrest of viral replication. Here, we
report on the synthesis of novel PI4KB inhibitors, which were rationally designed based on two
distinct structural types of inhibitors that bind in the ATP binding side of PI4KB. These
“hybrids” excel not only in outstanding inhibitory activity but they also show high selectivity to
PI4KB compared to other kinases. Thus, these compounds exert selective nanomolar or even
subnanomolar activity against PI4KB as well as profound antiviral effect against hepatitis C
virus, human rhinovirus and coxsackievirus B3. Our crystallographic analysis unveiled the exact
position of the side chains and explains their extensive contribution to the inhibitory activity.
INTRODUCTION
Positiveꢀsense single stranded RNA (+RNA) viruses replicate at specialized compartments
1
termed replication factories or membranous webs . These membraneꢀderived compartments are
2
specifically enriched in the phosphatidylinositol 4ꢀphosphate (PI4P) lipid . Humans have four
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 75
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
kinases that synthesize PI4P, two typical lipid kinases ꢀ PI4KA and PI4KB (also known as PI4K
IIIα and PI4K IIIβ) and two type II (or atypical) kinases ꢀ PI4K2A and PI4K2B (also known as
3
,4
PI4K IIα and PI4K IIβ) . In 2009 Berger and colleagues used siRNA screen to identify essential
host factors for the replication of hepatitis C virus (HCV) and found PI4KA to be an essential
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
host factor for HCV . Since then PI4KB has also been reported to be an essential host factor for
a plethora of +RNA viruses including SARS coronavirus, kobuvirus, poliovirus, coxsackievirus
6
ꢀ8
B3, and human rhinovirus . Surprisingly, so far PI4K2A or PI4K2B have never been reported
as a viral coꢀfactor. These enzymes are palmitoylated and can behave as integral transmembrane
proteins, thus, it might be too difficult for a virus to manipulate them.
Now it is generally accepted that most +RNA viruses hijack PI4K kinase to produce PI4Pꢀ
9
enriched membranes . Viruses have evolved different strategies of PI4K hijacking. The
10
aforementioned HCV uses its NS5A protein to directly hijack PI4KA . Many picornaviruses use
their nonꢀstructural 3A protein to hijack PI4KB. For instance, the 3A protein from the poliovirus,
coxsackievirus, rhinovirus or Aichi virus binds to the Golgi residing ACBD3 (AcylꢀCoA
8, 11, 12
Binding Domain Containing 3) protein, a known binding partner of the PI4KB
. However,
picornaviruses have also evolved additional means of PI4KB hijacking that are ACBD3
1
3ꢀ15
independent
.
PI4Ks are considered a potential pharmacological target because of their importance in the life
cycle of various many +RNA viruses. The discovery of their importance for +RNA viruses
aroused big interest in PI4Ks in the scientific community. During the last two years crystal
1
6ꢀ19
structures of P4KB, PI4K2A and PI4K2B have become available
providing critical
information for structure aided inhibitor design. The last missing piece is the structure of PI4KA
ACS Paragon Plus Environment
3

Page 5 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
20
perhaps due to difficulties in the expression of the recombinant protein . So far only a study
2
1
that used experiment guided homology modeling is available .
In the last few years, both academia and pharma invested substantial amounts of time in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
developing small molecule inhibitors targeting PI4KA and PI4KB. One of the first compounds, 1
2
2, 23
(PIK93), has attracted attention because of its selectivity towards PI4KB
, however, 1 has
2
4
also notable activity toward PI3 kinases . Later it was used as a starting point for developing
isoform specific inhibitors. A compound resembling 1 was reported to inhibit PI4KA with IC =
5
0
2
5
450 nM but the selectivity was rather poor (IC = 8000 nM for PI4KB) . Recently, very potent
5
0
inhibitors of PI4KA were reported – IC = ~1 nM for the best compound, however, the
5
0
2
6
selectivity was not optimal (IC = 250 nM for PI4KB) . Perhaps the best inhibitor of PI4KA
50
was reported by GlaxoSmithKleine which had an IC50 of 5nM against PI4KA and negligible
27
activity against PI4KB . We and others have also reported PI4KB inhibitors with varying
2
8ꢀ32
degrees of potency and specificity
with best compounds having an IC between 10−20 nM.
5
0
Important questions regarding the drugablity of these enzymes remain as knockout of these
2
5, 33
enzymes is detrimental to the animals
. However, upregulated PI4K activity is needed for a
successful viral infection. Total long time inhibition of either PI4KA or PI4KB might be toxic
but a short time dose (few days or a week) that would inhibit the enzymes only partially might be
well tolerated and still fatal for the virus. This is especially the case for PI4KB because PI4K2A
3
4, 35
can take over many of its roles at the Golgi
. Further studies are needed to decide whether
PI4KB is a valid pharmacological target and these studies will need very potent and extremely
specific inhibitors.
In this study we present the design and synthesis of novel PI4KB inhibitors, which exert high
inhibitory potency and outstanding selectivity in comparison to related lipid kinases.
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 75
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
Furthermore, our prototypical compounds show absolutely no significant inhibition of any other
kinases at 1 ꢀM concentration in kinomeꢀwide screen (Figure 3).
RESULTS AND DISCUSION
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
Design of inhibitors
Generally, specific PI4KB inhibitors targeting ATP binding side can be divided into two
structural groups. Firstly, several compounds derived from 1 exert both significant affinity and
high selectivity towards this enzyme. Secondly, derivatives with bicyclic central core related to 2
3
6
(
Tꢀ00127ꢀHEV1) were also identified as potent and selective inhibitors of PI4KB.
Figure 1. Structures of PI4KB inhibitors.
1
6, 28
Structural studies reported by us and Burke et al.
proved that both these types of compounds
bind in a similar manner in the ATP binding side and their structures can be easily overlaid and,
in principle, merged. This observation, together with subsequent docking studies, led to the
design of this second generation of PI4KB inhibitor. These newly designed inhibitors combine
features of both these aforementioned compounds into a single molecule. Originally, these
ACS Paragon Plus Environment
5

Page 7 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds were designed based on simple fusion of 1 and our compound 3 (MI14) (Figure 1
2
8
and Figure 2), which exerted a unique selectivity profile in our previous studies . Subsequently,
we tried to optimize the sidechains on both combinatorial as well as rational bases.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Our previous crystal structures revealed a lateral pocket, which is occupied by the amino sideꢀ
chain of the molecule (Figure 2A, in red). Evidently, this pocket can be accommodated by
significantly more bulky substituent than the (2ꢀacetamidoethyl)amino sideꢀchain employed in
our original study.
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 75
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 2. Design of new series of PI4KB inhibitors. A) Overlay of the crystal structure of PI4KB
with 1 (yellow) and the docking model of the preferred conformation of 3 (magenta) unveils two
sites of possible modifications. The red region can easily accommodate a large aromatic
substituent. The blue region appears open for further optimization of the aromatic part of the
molecule. BꢀD) Docking poses of derivatives 4, 5, and 8, respectively. The figure was prepared
using PyMOL.
ACS Paragon Plus Environment
7

Page 9 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Therefore, we investigated the effect of several related aromatic moieties that gave the best
results in our docking experiments (Table 1 and Figure 2). Interestingly, our docking results
suggest that compounds with various amino sideꢀchains can exploit different sites of the binding
canyon. In particular, the derivatives bearing the (2ꢀacetamidoethyl)amino sideꢀchain, e.g. 3, and
compounds with pyridine attached at position 3 tend to form an interaction with Asn 600
whereas the analogs attached at position 4 bind to Tyr 385 on the opposite side of the cleft.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Due to the significant effect on selectivity, the aromatic sidechain (Figure 2A, in blue) was more
difficult to target by means of rational design and therefore this part of the molecule was
elaborated by a combinatorial approach.
Inhibition of phosphatidylinositol 4-kinases
Initially we have focused our attention on the PI4KB potency and selectivity towards specific
PI4K isoforms. The enzymatic activity was determined using a luminescent ADPꢀGlo kinase
3
7
assay .
Firstly, we determined the inhibitory activity of derivatives with the optimized “amine sideꢀ
chain” of the molecule and it was encouraging to see similar trends in both the activity predicted
3
8
by our docking experiments performed in AutoDock Vina software and observed inhibition
constants in enzymatic assay. The data obtained for derivatives with 3ꢀ(N,Nꢀ
dimethylsulfamoyl)phenyl “aromatic sideꢀchain” are summarized in the Table 1. This part of the
study shows that the position of the nitrogen atom on the phenyl ring can significantly contribute
to the efficiency of the inhibition, which significantly correlated with results obtained by the
docking experiments. Thus the derivative connected to the central core at position 4 of the
pyridine ring (compound 4) exerted significantly higher potency in comparison with its
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 75
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
counterparts attached at position 2 or 3 – compounds 6 and 5, respectively. Subsequently, we
also tried to optimize the substituent on the pyrimidine ring. Since there is significant space next
to the pyrimidine core we gradually introduced methyl and ethyl group at the position 2 of the
pyrimidine of the side chain, which seemed to fit the best into this cavity of the enzyme based on
the docking experiments. Derivative 8 exerted significantly elevated inhibitory activity in
comparison to 3 with retention of excellent selectivity towards PI4KB with respect to the most
similar isoenzyme, PI4KA.
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
Therefore, this motive was used in our subsequent studies focused on the “aromatic sideꢀchain”
of the molecule. Here, we focused primarily on the additional effect of introducing the
substituent vicinal to the sulfonamide moiety and character of the sulfonamide moiety itself
(both based on the 1/3 hybrid design) (Table 2, compounds 8 and 9).
The installation of the sulfonamide moiety equivalent to 1 presented immediate success in
enhancing the inhibitory activity and compound 10 exerted significant activity while still
retaining the excellent selectivity. This series (Table 2, series A) gave a number of interesting
compounds with single digit nanomolar inhibitory activity and showed that the substitution of
the sulfonamide moiety has a significant impact on the selectivity towards PI4KB as well. The
most active compounds of the series were derivatives 10, 17 and 22, which all exerted low
nanomolar activity against PI4KB with outstanding selectivity. All these compounds shared the
same sulfonamide substituent – a free hydroxy group.
Series B of this study was mostly driven by our curiosity. Besides the derivatives with the
optimized “amino sideꢀchain”, we prepared derivatives with the original 2ꢀacetamidoethylamino
substituent as well in order to see how substantial is the impact of the more efficient but less
ACS Paragon Plus Environment
9

Page 11 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
flexible pyridine sideꢀchain. To our surprise, some of the analogs within the series provided
extremely high inhibitory activity against PI4KB with an outstanding PI4KB/PI4KA selectivity
ratio with the most active compound being derivative 25, which exerts subnanomolar activity
against PI4KB (IC = 0.98 nM) with PI4KA IC = 48 000 nM.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
50
Table 1. Correlation of the inhibitory potencies of the derivatives with modified “amino side
chain” with the results obtained from the docking experiments using AutoDock Vina Software.
Compd R
PI4KB
AutoDock
a
IC (nM) Vina
5
0
Score
9.9
4
26.4
5
6
7
74.6
145.4
33.7
9.8
9.7
N
10.2
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 75
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
8
12.5
54.0
10.4
3
9.3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Determined by three independent experiments.
Table 2.
Compd
Ar
Series PI4KΒ
IC (nM)
PI4KΑ
PI4K2A
% of control
(IC50 (nM))
ND
a
% of control
5
0
(
IC (nM))
50
8
A
A
12.5
ND
9
444.1
102
101
ACS Paragon Plus Environment
11

Page 13 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
0
1
A
A
A
6.1
42 (13 400)
9 (553)
70
106
101
97
O
S
O
N
O
OH
H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
15.7
17.6
12
13
A
59.2
80
99
14
15
16
A
A
A
9.0
14 (1 128)
100
104
112
O
S
S
O
O
N
O
O
N
O
H
17.2
19.1
50
77
O
N
H
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 75
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
7
A
6.2
58
109
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
8
9
A
A
8.6
58
74
105
106
1
27.3
20
A
41.5
61
115
21
A
A
B
56.0
5.3
84
59
63
114
104
112
2
2
3
2
22.0
O
S
O
N
O
OH
H
ACS Paragon Plus Environment
13

Page 15 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
4
5
B
B
26.9
0.98
13 (660)
102
88
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
93 (48 000)
O
S
N
O
O
O
2
6
B
189.4
62
60
27
28
29
B
B
B
93.5
32 (4 860)
105
107
83
O
S
S
O
O
N
O
O
N
O
H
134.8
163.3
80
98
O
N
H
30
B
43.8
72
88
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 75
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
3
1
2
B
B
19.9
2.4
77
72
93
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
104
33
34
B
14.9
43 (8 190)
85
B
B
B
105.8
7.23
80
108
77
3
5
6
34 (12 280)
44 (17 070)
3
146.9
73
O
S
H
O
N
O
N
H
3
B
54
(>100 000)
(>100 000)
a
Determined by three independent experiments.
Antiviral screening.
ACS Paragon Plus Environment
15

Page 17 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The obtained compounds were also screened against selected +RNA viruses ꢀ human rhinovirus
1
(HRV1), Coxsackie virus B3 (CVB3), hepatitis C virus (HCV ꢀ 1b and 2a genotype) and
Middle East respiratory syndrome coronavirus (MERSꢀCoV).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For clarity we present only the results for the compounds that had an IC lower than 20 nM in
5
0
the enzymatic assays (Table 2, additional data can be found in Supplementary Information
Table 2 and 3).
Antiviral screening unveiled a significant difference in activity of the A and B series. Generally,
the members of the A series exerted extremely high activity against HRV1, which in most cases
reached the detection limit of the assay (EC lower than 100 nM for all presented compounds).
5
0
Furthermore, most of the members of this series exerted significant inhibitory effect towards
CVB3 and HCV 1b. It is noteworthy, that derivatives 10, 17, and 22 which were the most active
members of this series in the enzymatic assay also exerted steady nanomolar activities in these
three antiviral assays. Notably, the effect on HCV 2a is rather negligible, which is consistent
3
9
with genotype dependent insensitivity of HCV to knockdown of PI4KB . In this respect, the
inactivity can be regarded as proof of selectivity towards PI4KB, as HCV 2a is significantly
more susceptible to inhibition of PI4KA.
In contrast, the compounds from series B exerted significantly lower nominal activity in cellꢀ
based assays. However, their toxicities were shifted in the similar fashion suggesting that the
compounds suffered from limited permeation through the cell membranes and therefore cannot
reach the side of the action. However, compound 25, the derivative exerting the highest activity
in the enzymatic assay, was potent enough to inhibit the sensitive viruses at the submicromolar
or low micromolar level.
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 75
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
None of the compounds exerted significant antiviral effect against MERSꢀCoV, which seems to
be in contrast with the reported implication of PI4KB in the replication of pathogenic
6
coronaviruses such as Severe acute respiratory syndrome coronavirus (SARSꢀCoV) .
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 3.
Compd Series HRV1
EC (nM)
CVB3
HCV 1b
HCV 2a
HeLa
EC (nM)
EC (nM)
EC (nM)
CC (nM)
50
50
50
50
50
8
1
1
1
1
1
1
1
1
2
2
3
3
3
3
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
62
1100
58
393
29
8689
15354
18333
5757
0
1
2
4
5
6
7
8
2
5
1
2
3
5
75
14402
26540
25988
8506
69
56
9196
1025
350
416
954
20
<10
<10
<10
<10
<10
<10
<10
655
990
716
5421
6756
83
25729
5558
151
<20
138
17
27940
>44444
20400
20147
9439
9144
61931
8399
12
667
<20
1033
2681
594
1826
3415
6853
476
38
17850
89223
445413
84176
131179
>50000
> 44444
35923
> 44444
> 44444
> 44444
5310
1320
5000
1661
Inhibition of other kinases by compound 10
ACS Paragon Plus Environment
17

Page 19 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Kinomeꢀwide profiling of compound 10 was performed by Carna Biosciences, Inc. This screen
unveiled a striking selectivity towards PI4KB since activity of 318 tested protein kinases was not
inhibited at 1 ꢀM concentration of compound 10 to more than to 50% of the control (Figure 3A).
Additionally, this compound was also screened against a panel of lipid kinases (PI4KA, PI4K2A,
PI3Kα, PI3Kβ, PI3Kδ, PI3Kγ) and none of these lipid kinases was inhibited with an IC lower
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
than 10 ꢀM (Figure 3B).
Figure 3. Graphical representation of the kinomeꢀwide profiling of compound 10. The tested
kinases are depicted in yellow. Only PI4KB was significantly inhibited at 1 ꢀM concentration.
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 75
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
Chemistry.
Synthesis of the target compounds started by the preparation of iodinated imidazo[1,2ꢀ
b]pyridazine derivatives with the appropriate amine side chain (Scheme 1).
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
Scheme 1. Preparation of iodinated precursors 37−42 by selective nucleophilic displacement of
a
one chlorine atom.
a
Reagents and conditions: (a) RꢀH, DIPEA, EtOH, heating.
ACS Paragon Plus Environment
19

Page 21 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
First, we substituted the original aminoethylacetamide side chain with positional isomers of
pyridinylmethanamine by means of nucleophilic aromatic substitution (EtOH, DIPEA, 70 °C) of
halogenated imidazo[1,2ꢀb]pyridazine derivatives (Scheme 1) in moderate to quantitative yields.
In consideration of the enzymatic binding site, the simple pyridine moiety was replaced by either
2ꢀmethyl or 2ꢀethylpyridine, which achieved better scores in our docking studies and filled up
the cavity of PI4KB with higher efficiency than unsubstituted the pyridines (see above). Both 2ꢀ
methyl and 2ꢀethylpyridinylmethanamines were prepared from 2ꢀchloroisonicotinonitrile in two
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
0
steps following the procedure of Ceccarelli et al.
a
Scheme 2. Preparation of derivatives 4−8 utilizing crossꢀcoupling.
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 75
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
R
N
R
N
N
a
N
N
N
Cl
Cl
I
O
S
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
N
O
3
7-41
4-8
Starting
compnd
Compnd
R
Yield
86%
N
N
H
N
4
37
38
39
40
41
H
N
5
6
7
8
80%
80%
80%
83%
H
N
N
N
N
H
N
H
N
a
Reagents and conditions: (a) ArꢀB(OH) , Pd(PPh ) , K CO , dioxane/H O (4:1), 95 °C, 16 h.
2
3 4
2
3
2
Scheme 3. Synthesis of derivatives 9 and 43 by crossꢀcoupling reactions with corresponding
a
boronic acids.
ACS Paragon Plus Environment
21

Page 23 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) ArꢀB(OH) , Pd(PPh ) , K CO , dioxane/H O (4:1), 95 °C, 16 h.
2
3 4
2
3
2
This series of iodo derivatives served as a modular toolbox for further exploration of SAR
focused on the aromatic side chain. Initially, we selected the aryl moiety that displayed the best
biological activity among our first series of PI4KB inhibitors ꢀ the 3ꢀ(N,Nꢀ
dimethylsulfamoyl)phenyl. Within this part of the study we also prepared derivatives with
various other aromatic substituents identified as potentially interesting in our previous work. The
majority of these derivatives were prepared by simple Suzuki crossꢀcoupling of six
iododerivatives with the appropriate boronic acid or its pinacol ester; the results are summarized
in the Supplementary information.
Derivatives 9 and 43 prepared by the Suzuki coupling procedure were chlorosulfonylated (0
°
C, CH Cl ) to provide derivatives 44 and 45 in quantitative yield. Crude derivatives 44 and 45
2 2
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 75
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
were treated with various primary and secondary, cyclic or acyclic amines (DIPEA, CH Cl , rt)
2
2
providing pꢀmethoxysulfoamides in moderate to good yields. Results are summarized in Scheme
4.
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
Scheme 4. Preparation of sulfonamide derivatives 10−36.
ACS Paragon Plus Environment
23

Page 25 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Reagents and conditions: (a) chlorosulfonic acid, rt, 16 h, quant.; (b) amine (XꢀH), DIPEA,
CH CL , rt, 16 h.
2
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
Binding mode of novel inhibitors – structural analysis
We used protein crystallography to gain insight in the binding mode of our new inhibitors.
41
PI4KB contains unstructured regions that impede crystallization . Thus we adopted the strategy
16
introduced by Burke et al. – we removed the disordered loops and crystallized PI4KB together
with Rab11 (wt or Q70L mutant) (Figure 4, Supplementary information Figure S1). Using this
strategy, we were able to obtain five crystal structures with members of Series B (compounds 23,
2
4, 25, 33, and 35) (Table 4). Generally, our crystallographic analysis was in very good
agreement with our docking studies. It revealed that the binding mode of the ligand can be
severely influenced by the sulfonamide substituent. Although water molecules probably play an
important role in the interaction of both “amino sideꢀchain” and “aromatic sideꢀchain” we did not
model the water molecules due to limited resolution of our crystal structures. In particular, the
position of the “amino sideꢀchain”, although quite remote from the “aromatic side chain”, differs
significantly depending on the sulfonamide moiety attached. Thus, the (2ꢀacetamidoethyl)amino
sideꢀchain can directly form hydrogen bonds with Asn615 (ligand 25, Figure 4D), Tyr385
(
ligand 33, Figure 4E) or even with Ala616 (ligand 35, Figure 4F). However, it seems to be clear
that the interaction with these residues can be mediated also through a water bridge, which seems
to be the case for compounds 23 and 24 (Figure 4B and 4C, respectively). In addition, the
sulfonamide moiety can form hydrogen bonds with Gly675 and/or Asn676 as is obvious from the
crystal structures containing derivative 24 and 33 (Figure 4C and 4E, respectively). In contrast,
the interaction with these two residues can be also provided through water molecules as might be
ACS Paragon Plus Environment
25

Page 27 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the case for compound 25 (Figure 4D). An interesting difference was observed for structurally
related compounds 23 (Figure 4B) and 35 (Figure 4F). Compound 23 forms the interaction with
Lys584 through hydrogen bonding with the oxygen bonded on the sulfur atom whereas the
hydroxyl group on the rear of the sulfonamide sideꢀchain provides the interaction with the same
residue as in the case of compound 35. Therefore, it seems that the exact ligand’s position and
the efficiency of its interaction with the kinase is highly dependent on the sulfonamide
substituent and can be fineꢀtuned by careful selection of this moiety. We also hypothesize that
this effect is responsible for the fluctuation in the results observed for these diverse sulfonamide
moieties in both series we prepared.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Statistics of crystallographic data collection and refinement.
Crystal
PI4KIIIβ:wtRab1 PI4KIIIβ:wtRab1 PI4KIIIβ:Q70 PI4KIIIβ:Q70 PI4KIIIβ:Q70
1
+ 25
1 + 33
L Rab11 + 24 L Rab11 + 23 L Rab11 + 35
PDB code
5FBL
5FBQ
5FBR
5FBV
5FBW
Space group P 2 2 2
P 2 2 2
P 2 2 2
P 2 2 2
P 2 2 2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Cell
dimension
a = 49.5 Å, b = a = 48.8 Å, b = a = 49.3 Å, b a = 48.5 Å, b a = 49.1 Å, b
104.9 Å, c = 101.2 Å, c = = 103.9 Å, c = = 102.8 Å, c = = 104.1 Å, c
188.0 Å
188.5 Å
187.6 Å
186.1 Å
= 187.6 Å
Xꢀray source BESSY ID 14ꢀ1
BESSY ID 14ꢀ1
BESSY
14ꢀ1
ID BESSY
14ꢀ1
ID BESSY
14ꢀ1
ID
Wavelength, 0.9796
Å
0.9796
0.9796
0.9796
0.9796
Resolution,
Å
47.01 ꢀ 3.37 (3.49 48.85
ꢀ 3.79 47.65 ꢀ 3.28 46.97 ꢀ 3.28 47.47 ꢀ 3.48
ꢀ 3.37)
(3.92 ꢀ 3.78)
(3.40 ꢀ 3.28)
(3.40 ꢀ 3.28)
(3.61 ꢀ 3.48)
No.
of 14419 (1382)
9508 (865)
15037 (1135)
14745 (1381)
12696 (1166)
unique
reflections
I/σ (I)
14.11 (1.71)
8.33 (1.45)
3.85
12.75 (1.35)
3.32
7.63 (0.95)
3.45
6.77 (0.87)
3.67
I/σ (I) = 2 at 3.43
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 75
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
resolution, Å
Rmeans, %
11.3
21.1
12.8
23.0
19.1
Data
completeness
99.54 (97.19)
96.34 (91.53)
97.42 (76.48)
99.01 (95.11)
98.32 (93.65)
,
%
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
Multiplicity
Rwork, %
3.85
5.40
6.33
3.44
4.27
24.28 (35.68) 25.18 (35.94)
26.88 (34.96) 30.44 (38.12)
22.61 (35.26)
26.38 (38.44)
0.004
23.71 (39.66)
28.14 (43.98)
0.012
24.62 (40.30)
28.31 (42.35)
0.005
R , %
free
rms bond angle 0.004
deviation, (deg)
0.003
rms bond angle 1.01
deviation, Å
0.89
0.91
1.54
1.15
Ramachandran
outliers/
favored)
0% / 96%
0% / 97%
0% / 95%
0% / 96%
0% / 96%
(
ACS Paragon Plus Environment
27

Page 29 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Structural analysis of newly the prepared inhibitors. A) Structure of the whole
crystalized complex of PI4KB with Rab11. BꢀF) Binding modes of inhibitors 23, 24, 25, 33, and
35. The figure was prepared using PyMOL.
ACS Paragon Plus Environment
28