SGC-GAK-1: A Chemical Probe for Cyclin G Associated Kinase (GAK)

Article abstract of DOI:10.1021/acs.jmedchem.8b01213

We describe SGC-GAK-1 (11), a potent, selective, and cell-active inhibitor of cyclin G-associated kinase (GAK), together with a structurally related negative control SGC-GAK-1N (14). 11 was highly selective in an in vitro kinome-wide screen, but cellular

Full text of DOI:10.1021/acs.jmedchem.8b01213

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Brief Article
SGC-GAK-1: a chemical probe for cyclin G associated kinase (GAK)
Christopher R. M. Asquith, Benedict-Tilman Berger, Jing Wan, James M. Bennett, Stephen
Joseph Capuzzi, Daniel J. Crona, David Drewry, Michael P. East, Jonathan M Elkins,
Oleg Fedorov, Paulo H. Godoi, Debra M. Hunter, Stefan Knapp, Susanne Müller, Chad
D. Torrice, Carrow Iona Wells, Henry Shelton Earp, Tim Willson, and William Zuercher
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01213 • Publication Date (Web): 15 Feb 2019
Downloaded from http://pubs.acs.org on February 17, 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 10
Journal of Medicinal Chemistry
1
2
3
4
5
6
Willson, Tim; University of North Carolina at Chapel Hill Eshelman School
of Pharmacy, Structural Genomics Consortium
Zuercher, William; University of North Carolina at Chapel Hill Eshelman
School of Pharmacy, Structural Genomics Consortium; University of
North Carolina Lineberger Comprehensive Cancer Center
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 10
1
2
3
4
5
6
7
8
9
SGC-GAK-1: a chemical probe for cyclin G associated kinase (GAK)
Christopher R. M. Asquith,† Benedict-Tilman Berger,‡, Jing Wan,§,^ James M. Bennett, Stephen
J. Capuzzi, Daniel J. Crona,,^ David H. Drewry,† Michael P. East,‡‡ Jonathan M. Elkins,, Oleg
Fedorov, Paulo H. Godoi, Debra M. Hunter,§,^ Stefan Knapp,‡, Susanne Müller,‡ Chad D.
Torrice, Carrow I. Wells,† H. Shelton Earp,§,^ Timothy M. Willson,† and William J. Zuercher†,^,*
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
†Structural Genomics Consortium, UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599, US
‡Structural Genomics Consortium, Johann Wolfgang Goethe University, Buchmann Institute for Molecular Life
Sciences, Max-von-Laue-Str. 15, D-60438 Frankfurt am Main, DE
Institute for Pharmaceutical Chemistry, Johann Wolfgang Goethe University, Max-von-Laue-Str. 9, D-60438
Frankfurt am Main, DE
§Department of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, US
^Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, NC 27599, US
Structural Genomics Consortium and Target Discovery Institute, Nuffield Department of Clinical Medicine,
University of Oxford, Old Road Campus Research Building, Oxford, OX3 7DQ, UK
Division of Chemical Biology and Medicinal Chemistry, UNC Eshelman School of Pharmacy, University of North
Carolina at Chapel Hill, Chapel Hill, NC 27599, US
Division of Pharmacotherapy and Experimental Therapeutics, UNC Eshelman School of Pharmacy, University of
North Carolina at Chapel Hill, Chapel Hill, NC 27599, US
‡‡Department of Pharmacology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, US
Structural Genomics Consortium, Universidade Estadual de Campinas, Campinas, São Paulo, 13083-886, BR
ABSTRACT: We describe SGC-GAK-1 (11), a potent, selective, and cell-active inhibitor of cyclin G-associated kinase
(GAK), together with a structurally-related negative control SGC-GAK-1N (14). 11 was highly selective in an in vitro
kinome-wide screen, but cellular engagement assays defined RIPK2 as a collateral target. We identified 18 as a potent
RIPK2 inhibitor lacking GAK activity. Together, this chemical probe set can be used to interrogate GAK cellular biology.
in membrane trafficking and sorting of proteins and as an
essential cofactor for HSC70-dependent uncoating of
clathrin-coated vesicles in the cytoplasm.6, 7 GAK is also
required for maintenance of centrosome maturation and
progression through mitosis.8 GAK is over-expressed in
osteosarcoma cell lines and tissues where it contributes to
proliferation and survival.9 Notably, GAK expression
increases during prostate cancer progression to androgen
independence and is positively correlated with Gleason
score in resections from prostate cancer patients.10, 11
Introduction
Cyclin G-associated kinase (GAK) is
a
160 kDa
serine/threonine kinase originally identified as a direct
association partner of cyclin G.1 GAK is a member of the
numb-associated kinase (NAK) family, which includes
AAK1 (adaptor protein 2-associated kinase), STK16/MPSK1
(serine/threonine
palmitoylated
kinase
serine/threonine
16/myristoylated
kinase 1),
and
and
BMP2K/BIKE (BMP-2 inducible kinase).2 In addition to its
kinase domain, the C-terminus of GAK protein has high
homology with auxilin and tensin.3 GAK is ubiquitously
expressed and within the cell localizes to the Golgi
complex, cytoplasm, and nucleus.4
We have a program to develop small molecule chemical
probes to elucidate the biological role of kinases such as
GAK that belong the lesser-studied portion of the
kinome.12 To be useful research reagents, these chemical
probes must be potent, selective, and cell-active.13
Specifically the aspirational criteria we follow are (i) in
vitro biochemical IC50 < 50 nM; (ii) ≥ 30-fold selective
relative to other kinases in a large assay panel such as
Genetic studies have implicated GAK in several diverse
biological processes. Genome-wide association studies
have identified single nucleotide polymorphisms in gak
associated with susceptibility to Parkinson’s disease.5
GAK, like other members of the NAK family, is involved
ACS Paragon Plus Environment

Page 3 of 10
Journal of Medicinal Chemistry
DiscoverX KINOMEscan; and (iii) cellular activity or nonpolar substituents at the quinoline 6- and 7-positions
1
2
3
4
5
6
7
8
target engagement IC50 < 1 µM. In addition, the probe
should be accompanied by a structurally related negative
control that is at least several orders of magnitude less
potent at the primary kinase target.13
might show improved selectivity over RIPK2.
To investigate this hypothesis and define the SAR of
ADCK3 and NLK, we profiled several quinolines prepared
in our previous study against the three collateral targets
(Table 1).17 6,7-Dimethoxyquinoline 3 had the highest
GAK affinity but also showed increased binding affinity to
RIPK2 relative to 2. Replacement of either methoxy group
with a hydrogen, as in 4 and 5, led to reduced affinity on
all four kinases. Notably, the unsubstituted quinoline 6
showed an improved selectivity profile, although it had
the poorest affinity for GAK among the analogs. The
nitrile isomers 7 and 8 both retained single digit
nanomolar affinity for GAK yet had different effects on
selectivity profile: the 6-isomer 7 had an improved
selectivity profile while the 7-isomer 8 showed binding to
all three off-targets within ten-fold. A tert-butyl group in
the 6-position (9) led to improvement of the selectivity
profile against RIPK2 and ADCK3 with a decrease in
selectivity over NLK compared with 2.
A number of quinazoline-based kinase inhibitors show
cross-reactivity with GAK, including the approved drugs
gefitinib and erlotinib (Figure 1).14 These drugs were
designed as inhibitors of either epidermal growth factor
receptor (EGFR) or SRC kinase and show similar or higher
affinity for several other kinases, making them ineffective
tools for studying GAK biology. Thus, it is not known
whether the efficacy or adverse events observed with
these kinase inhibitor drugs are associated with their
cross-activity on GAK. Notably, it has been proposed that
GAK inhibition causes toxicity due to pulmonary alveolar
dysfunction. This hypothesis is based in part on the
phenotype observed in transgenic mice expressing
catalytically inactive GAK but has not yet been tested
with a selective small molecule GAK inhibitor.15
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
F
Table 1. Kinase affinity of 4-anilinoquinolines.
NH
Cl
NH
N
O
O
O
O
N
N
OMe
OMe
N
N
OMe
Gefitinib, GAK KD = 13 nM
Erlotinib, GAK KD = 3.1 nM
OMe
MeO
R1
R2
GAKa
31
RIPK2 a ADCK3 a NLK a
OMe
MeO
2
3
CF3
H
180
42
220
250
1200
2000
4100
170
790
410
N
MeO
S
MeO
NH
OMe OMe
2.9
17
N
CF3
N
4
5
6
7
8
9
H
OMe
H
OMe
H
200
170
2900
88
700
2600
6200
29
N
O
1, GAK KD = 8.9 nM
2, GAK KD = 31 nM
21
H
59
Figure 1. Previously described inhibitors of GAK.
CN
H
H
3.9
9.7
25
Isothiazolo[5,4-b]pyridine 1 has been described as a
selective inhibitor of GAK but shows cross-reactivity with
other kinases, including KIT, PDGFRB, FLT3, and MEK5,16
any of which could lead to confounding biology in cells.
The availability of a GAK chemical probe with improved
selectivity would be a valuable tool to probe GAK biology.
We previously described 4-anilinoquinoline inhibitors of
GAK, exemplified by 2.17 The optimization of this series to
a GAK chemical probe is described.
CN
H
96
66
28
t-Bu
670
740
350
aKD (nM) in competition binding assay (DiscoverX, n = 2).
Comparison of the two matched molecular pairs with
methoxy and nitrile (4 and 8; 5 and 7 respectively)
suggested that electron withdrawing substituents at the
quinoline 6-position showed improved selectivity and
warranted further investigation. Quinolines 10-12 were
Results
prepared
by
nucleophilic
substitution
of
4-
Quinoline 2 was profiled at 1 µM across a panel of over
400 wild-type human kinases using heterogeneous assays
employing displacement of kinase protein from an
immobilized inhibitor. Subsequent KD determination for
those kinases with > 50% binding at 1 µM identified three
kinases (receptor-interacting protein kinase 2, RIPK2;
AarF domain containing kinase 3, ADCK3; and nemo-like
kinase, NLK) with KD values within 30-fold of that of GAK
(Table S1). While little is known about the activity of
quinazolines on ADCK3 and NLK, a prior study of
quinazoline inhibitors of RIPK2 showed H-bonding
between Ser25 of RIPK2 near the solvent-exposed portion
of the ATP-binding pocket.18 Since GAK has Ala47 as the
corresponding residue, we hypothesized that small,
chloroquinolines with 3,4,5-trimethoxyaniline.17 The
halogen-substituted analogs 10-12 showed high GAK
affinity, all with KD < 10 nM (Table 2). 11 also displayed
remarkable selectivity, with the closest off-target kinase
(RIPK2) having > 50-fold weaker affinity.
To confirm the selectivity profiles across the NAK
subfamily, the affinity of 11 was evaluated in
homogeneous time-resolved fluorescence resonance
energy transfer (TR-FRET)-based ligand binding
displacement assays (Figure 2A). 11 displayed an
impressive profile of over 16000-fold selectivity relative to
the other three NAK family members. In addition, a
differential scanning fluorimetry (DSF) assay19 and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
isothermal titration calorimetry (ITC) were employed to
Page 4 of 10
1
2
3
4
5
6
7
8
confirm direct interaction with GAK (Figure 2B and 2C).
11 increased the thermal stability of the isolated GAK
kinase domain, suggesting strong binding of the inhibitor.
ITC revealed potent affinity for GAK (KD = 4.5 nM). We
assessed the kinome selectivity of 11 in the KINOMEscan
assay panel (Figure 2D and Table S2). KD values were
measured for kinases with binding greater than 50% at 1
µM (Table S3). 11 showed > 50-fold selectivity relative to
all 400 kinases. The closest off-targets were RIPK2,
ADCK3, NLK and ACVR1.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Kinase binding of halogenated quiniolines.
R
Cl
Br
I
GAK a
6.7
RIPK2a
85
ADCK3a
170
NLKa
650
10
11
1.9
110
190
520
Figure 2. (A) Binding affinity of 11 for NAK family kinases.
12
7.9
69
190
660
aTemperature shift of GAK denaturation (n = 4); bTR-
FRET-based ligand binding displacement assay (n = 1). (B)
DSF curve of GAK and 11; (C) ITC of GAK and 11; (D)
KINOMEscan results.
aKD (nM), competition binding assay (DiscoverX, n = 2).
An established best practice for the use of chemical
probes in biological assays is to test in parallel a
compound closely related in structure to the probe but
inactive for the target protein.13 Demonstration of
divergent results between the chemical probe and its
negative control increases confidence that the
engagement of the target protein is responsible for the
observed biology. We designed and prepared two
potential negative controls that were anticipated to lack
GAK activity. First, 13 incorporated a 2-methylquinoline
that would directly hinder hydrogen bonding to the hinge
region of the kinase. Second, the N-methyl 14 was
expected to disrupt a water network necessary for
favorable compound binding (Asquith, C. R. M. and
Laitinen, T., unpublished results). Neither compound
induced thermal stabilization of GAK protein, and both 13
and 14 demonstrated >10,000-fold weaker GAK binding
affinity in the TR-FRET assay (Table 3). KINOMEscan
and subsequent KD determinations of 13 and 14 identified
only a single kinase with < 1 µM affinity (14: interleukin 1
receptor associated kinase 3, IRAK3 KD = 740 nM) (Tables
S2 and S4). Both compounds were suitable as GAK
negative controls, and we selected 14 based on (a) its
higher selectivity in the NAK family ligand binding
displacement assay and (b) the structural modification
leading to loss of GAK activity was not expected to
directly interfere with kinase hinge binding.
Table 3. NAK family binding of 13 and 14.
GAKa
-0.52
-0.30
AAK1 b BMP2K b GAK b
STK16 b
>100
15
13
14
>100
78
>100
95
19
2.5
aΔTm (°C) measured by DSF; bKi (µM) in TR-FRET-based
ligand binding displacement assay (n = 1).
absence of compound, the tracer molecule and fusion
protein were in proximity and able to generate an
observable bioluminescence resonance energy transfer
(BRET) signal. Increasing concentrations of compound
displaced the tracer, causing a loss of BRET signal. 11 had
high affinity for GAK in cells (IC50 = 120 nM), although the
measured IC50 was 63-fold higher than the in vitro KD
value. We interpret the difference as most likely due to
competition with ATP in cells. The negative control 14
showed no cellular binding to GAK up to 5 µM.
Notably, the previously reported GAK inhibitor 1 had
micromolar GAK cellular potency, approximately ten-fold
weaker than 11. We also evaluated 11 and 14 in a RIPK2
NanoBRET assay as this kinase was the closest off-target.
We were surprised that 11 had only three-fold selectivity
for GAK over RIPK2 in cells (Figure 3, GAK IC50 =120 ±
50 nM; RIPK2 IC50 = 360 ± 190 nM). The negative control
14 and the previously described GAK inhibitor 1 showed
no affinity for RIPK2 up to 10 µM.
Having confirmed potent GAK binding and selectivity
relative to other NAK family members, we next sought to
demonstrate that activity and selectivity of 11 in live cells
using a NanoBRET target engagement assay.20 11 and 14
were evaluated for their ability to displace a fluorescent
tracer molecule from the ATP binding site of a transiently
expressed full length GAK protein N-terminally fused
with NanoLuc luciferase (Nluc) in HEK293T cells. In the
ACS Paragon Plus Environment

Page 5 of 10
Journal of Medicinal Chemistry
growing VCaP cells, preventing the determination of an
1
2
3
4
5
6
7
accurate IC50 value. Overall, these results demonstrate
robust effects of 11 in a subset of prostate cancer cell lines.
Table 5. Prostate cancer cell line viability at 72 h.
Compound
LNCaPa
0.65  0.15
> 100
22Rv1a
0.17  0.05
>100
11
14
18
Figure 3. Cellular target engagement assays. a. GAK, b.
RIPK2. Cntrl: DMSO + tracer. Background: DMSO, no tracer.
8
9
> 100
99  26
Given the observed cross-activity of 11 with RIPK2 in cells,
we sought to identify an additional control with RIPK2
but not GAK activity. Surveying known RIPK2 inhibitors
(Table 4 and Figure S1), we first evaluated OD36 (15),
since it had previously been characterized as selective for
RIPK2 across a panel of over 400 wild type human kinase
assays (although this panel did not include GAK).21 We
also expected the distinct macrocyclic chemotype of 15
might have a lower propensity for binding GAK relative to
a quinoline and were surprised to observe that 15 was
nearly equipotent with RIPK2 and GAK in cells. Affinity of
15 for GAK was confirmed in an in vitro binding assay.
Next, we selected three quinazoline-based RIPK2
inhibitors (16-18) reported by GlaxoSmithKline.18,22 16 and
17 displayed high GAK affinity and were not subsequently
evaluated for cellular target engagement. 18 had poor
affinity for GAK, both biochemically and in cells (Figure
3).23 18 demonstrated high RIPK2 cellular affinity and was
selected as RIPK2 control.
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
aIC50  SEM (µM), (n = 3).
Previous reports have demonstrated that GAK is required
for proper mitotic progression.24 Specifically, siRNA
knockdown of GAK in HeLa cells induced cell cycle arrest
at the metaphase and increase levels of phosphorylated
histone H3 Ser10.24 Prolonged metaphase arrest is known
to lead to activation of apoptotic pathways.25 We observed
that treatment of 22Rv1 cells with 11 led to PARP cleavage,
a marker of cells undergoing apoptosis (Figure S3), and
an increase in phosphorylated histone H3 Ser10, as had
been observed in with GAK targeting siRNA.24 Thus, our
probe set demonstrated that pharmacological inhibition
of the GAK kinase domain can phenocopy the effect on
cell growth shown by depletion of GAK protein.
Discussion and Conclusions
We describe the identification of a potent, selective, and
cell-active GAK probe, SGC-GAK-1 (11), its negative
control SGC-GAK-1N (14), and the RIPK2 control 18. 11
was characterized for selectivity across a broad panel of
kinase assays. The use of controls 14 and 18 can further
increase confidence in GAK inhibition as the mediator of
phenotypes induced by 11 in cells. SGC-GAK-1 (11) has
several advantages over the previously reported
isothiazolo[4,3-b]pyridine GAK inhibitor 1, including
superior cellular potency and improved kinase selectivity.
The GAK probe set (11, 14, 18) has been made openly
Table 4. Evaluation of potential RIPK2 controls.
Compound
15 (OD36)
16 (GSK583)
17
GAK a
0.0038
0.22
GAK b
0.060
nt
RIPK2b
0.011
nt
0.049
8.2
nt
nt
18
> 10
0.0020
a TR-FRET ligand binding displacement assay Ki (µM) (n
available
(see
https://www.thesgc.org/chemical-
= 2); b NanoBRET assay IC50 (nM) (n = 2, nt = not tested).
probes/SGC-GAK-1). We expect it will find use in
elucidating GAK biology.
We propose that SGC-GAK-1 (11) and the controls SGC-
GAK-1N (14) and 18 can be used in parallel as a probe set
to study GAK biology in cells. As an illustrative example,
we used the probe set to explore the potential role of GAK
in prostate cancer.11 The probe set was tested for effects
on the viability of five well-characterized prostate cancer
cell lines (PC3, DU145, LNCaP, VCaP, and 22Rv1 cells).
Western blotting experiments confirmed that GAK was
expressed at similar levels in these cell lines (Figure S2).
In an initial experiment, the cell lines were incubated
with 11, 14, and 18 at three different concentrations (0.1, 1,
and 10 µM), and viability was measured at 48 h or 72 h.
No effect was observed with the controls 14 or 18 in any of
the cell lines at these concentrations. However, SGC-
GAK-1 (11) showed strong growth inhibition in LNCaP,
VCaP, and 22Rv1 cells at 10 µM, but minimal effect in PC3
and DU145 cells. In dose response 11 showed potent
antiproliferative activity in LNCaP and 22Rv1 cells at a
similar dose range that it engages GAK (Table 5).
However, variable results were observed in the slow
The identification of RIPK2 as a target of 11 is surprising
as GAK and RIPK2 belong to distal branches of the
kinome and have divergent ATP-binding pockets—of the
28 residues lining the pockets, only 14 are identical or
similar. The propensity of 11 to bind both kinases results
from higher order interactions not presently understood.
The narrowed GAK/RIPK2 selectivity window in cells
highlights the importance of profiling in relevant assay
systems. 11 demonstrated a > 50-fold difference in affinity
between GAK and RIPK2 in an in vitro binding assay.
whereas it was within 4-fold affinity on the two kinases in
live cell target engagement assays involving full length
protein, physiological levels of ATP, and endogenous
protein partners.20 We identified the quinazoline 18 as a
RIPK2 inhibitor that lacks affinity for GAK as a means to
control for this off-target in cells. There remains a finite
possibility that cellular screening would reveal additional
potent kinase targets of 11, but the KINOMEscan data
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 10
MHz, DMSO-d6) δ 157.3, 154.7, 154.1 (s, 2C), 136.9, 133.2, 129.2-129.1
suggest this is highly unlikely. Protein kinase inhibitors
such as 11 also have the potential to interact with non-
kinase targets.26 The use of structurally related 14 and 18
is intended to control for this possibility. For these
reasons 11, 14, and 18 should be used in parallel to support
robust interpretation of the biology of GAK inhibition.
(m, 1C), 126.7, 126.7, 126.4, 125.7, 123.0, 122.4 (d, J = 3.8 Hz), 116.2,
103.4, 102.0, 60.6, 56.6 (s, 2C), 20.9. HRMS-ESI (m/z): [M+H]+
calcd for C20H19F3N2O3: 393.1426, found 393.1413; LC - Tr = 4.02
min, purity > 98 %.
1
2
3
4
Supporting Information
5
6
7
8
The Supporting Information is available free of charge on the
ACS Publications website at DOI: XXXXX. Compound
synthesis, assay protocols, tabulated assay output, dose-
response curves, and supporting figures and tables (PDF)
Molecular formula strings with in vitro activity data (CSV)
The observation that GAK expression is associated with
prostate cancer progression prompted us to explore the
effect of GAK inhibition across a range of cell lines. To
date studies of the role of GAK in prostate cancer have
utilized knock-down of the whole protein.27 We now
show that inhibition of the GAK kinase domain can block
the growth of prostate cancer cells. Our results
demonstrate that the kinase and not the auxilin-
homology domain is critical for this phenotype in cells.
SGC-GAK-1 (11), but not the controls 14 and 18, robustly
blocked growth of 22Rv1 and LNCaP cell lines. DU145 or
PC3 cell lines were less responsive. Since neither DU145
nor PC3 express the AR, their proliferation is not expected
to be AR-dependent. In contrast LNCaP and 22Rv1 are AR
positive and are cell models for androgen-dependent
growth. Interestingly, the most potently inhibited cell line
(22Rv1) expresses the AR-V7 and other splice variants
often increased in patients who become resistant to
androgen deprivation.28 Why the AR-V7 expressing lines
are more sensitive and what are the mechanisms by which
GAK inhibitors target a subset of AR expressing prostate
cancer cells are the subjects of ongoing investigation.
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
AUTHOR INFORMATION
Corresponding Author
*Phone: +1 919-962-5349. E-mail: william.zuercher@unc.edu.
ORCID
Christopher Asquith: 0000-0001-5871-3458
Benedict-Tilman Berger: 0000-0002-3314-2617
Stephen Capuzzi 0000-0003-0001-8211
David Drewry: 0000-0001-5973-5798
Jonathan Elkins: 0000-0003-2858-8929
Stefan Knapp: 0000-0001-5995-6494
Susanne Müller: 0000-0003-2402-4157
Carrow Wells: 0000-0003-4799-6792
Timothy Willson: 0000-0003-4181-8223
William Zuercher: 0000-0002-9836-0068
Author Contributions
The manuscript was written through contributions of all
authors. All authors approved the final manuscript version.
In summary, we have generated the first chemical probe
set for the poorly studied kinase GAK. SGC-GAK-1 (11),
when used in parallel with 14 and 18, can be used to study
the cell biology of GAK. We demonstrate that the GAK
kinase domain is a potential target for inhibition of
prostate cancer growth in cells that express AR splice
variants associated with poor clinical prognosis.
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
The SGC is a registered charity (number 1097737) that
receives funds from AbbVie, Bayer Pharma AG, Boehringer
Ingelheim, Canada Foundation for Innovation, Eshelman
Institute for Innovation, Genome Canada, Innovative
Medicines Initiative (EU/EFPIA) [ULTRA-DD grant no.
115766], Janssen, Merck KGaA Darmstadt Germany, MSD,
Novartis Pharma AG, Ontario Ministry of Economic
Development and Innovation, Pfizer, São Paulo Research
Experimental
2-14 were synthesized as previously reported, and 2-9 were
consistent with previous reports.17 General procedures and
characterization of 10, 12, and 13 are included in the SI. All
compounds were > 98 % pure by 1H/13C NMR and LCMS.
General
procedure
for
the
synthesis
of
4-
anilinoquinolines: 6-Bromo-4-chloroquinoline (1.0 eq.), 3,4,5-
trimethoxyaniline (1.1 eq.), and i-Pr2NEt (2.5 eq.) were suspended
in ethanol (30 mL) and refluxed for 18 h. The crude mixture was
purified by flash chromatography using EtOAc:hexane followed
by 1-5 % methanol in EtOAc. After solvent removal under
reduced pressure, the product was obtained.
Foundation-FAPESP,
Takeda,
and
Wellcome
[106169/ZZ14/Z]. Antti Poso and Tuomo Laitinen (University
of Eastern Finland) are thanked for informative discussions.
We also thank Dr. Brandie Ehrmann for LC-MS/HRMS
support provided by the Mass Spectrometry Core Laboratory
at the University of North Carolina at Chapel Hill.
6-Bromo-N-(3,4,5-trimethoxyphenyl)quinolin-4-amine
(11) was obtained as a mustard solid (310 mg, 0.79 mmol, 64%)
MP 178-180 °C; 1H NMR (400 MHz, DMSO-d6) δ 11.06 (s, 1H), 9.19
(d, J = 1.7 Hz, 1H), 8.48 (d, J = 6.9 Hz, 1H), 8.32 – 7.82 (m, 2H),
6.92 (d, J = 6.9 Hz, 1H), 6.82 (s, 2H), 3.80 (s, 6H), 3.72 (s, 3H); 13C
NMR (101 MHz, DMSO-d6) δ 154.0, 153.6 (s, 2C), 143.0, 137.6,
136.5, 136.3, 132.8, 126.2, 122.7, 119.7, 118.5, 103.0 (s, 2C), 100.9, 60.2,
56.1 (s, 2C). HRMS-ESI (m/z): [M+H]+ calcd for C18H18N2O3Br:
389.0500, found 389.0490; LC - Tr = 3.60 min, purity > 98 %.
2-Methyl-6-(trifluoromethyl)-N-(3,4,5-trimethoxyphenyl)-
quinolin-4-amine (14) was obtained as a yellow solid (330 mg,
ABBREVIATIONS USED
DSF, differential scanning fluorimetry; GAK, cyclin
associated kinase; ITC, isothermal titration calorimetry
G
REFERENCES
1. Kanaoka, Y.; Kimura, S. H.; Okazaki, I.; Ikeda, M.; Nojima,
H., GAK: a cyclin G associated kinase contains a tensin/auxilin-
like domain. FEBS Lett 1997, 402, 73-80.
2. Sorrell, F. J.; Szklarz, M.; Abdul Azeez, K. R.; Elkins, J. M.;
Knapp, S., Family-wide structural analysis of human numb-
associated protein kinases. Structure 2016, 24, 401-411.
1
0.84 mmol, 69%) MP 240 °C decomp.; H NMR (400 MHz,
DMSO-d6) δ 10.87 (s, 1H), 9.20 (s, 1H), 8.23 (s, 2H), 6.87 (s, 1H),
6.79 (s, 2H), 3.81 (s, 6H), 3.73 (s, 3H), 2.65 (s, 3H); 13C NMR (101
ACS Paragon Plus Environment

Page 7 of 10
Journal of Medicinal Chemistry
3. Kimura, S. H.; Tsuruga, H.; Yabuta, N.; Endo, Y.; Nojima, H.,
Structure, expression, and chromosomal localization of human
GAK. Genomics 1997, 44, 179-187.
4. Sato, J.; Shimizu, H.; Kasama, T.; Yabuta, N.; Nojima, H.,
GAK, a regulator of clathrin-mediated membrane trafficking,
localizes not only in the cytoplasm but also in the nucleus. Genes
Cells 2009, 14, 627-641.
gefitinib target GAK is caused by pulmonary dysfunction. PLoS
One 2011, 6, e26034.
1
2
3
4
16. Kovackova, S.; Chang, L.; Bekerman, E.; Neveu, G.;
Barouch-Bentov, R.; Chaikuad, A.; Heroven, C.; Sala, M.; De
Jonghe, S.; Knapp, S.; Einav, S.; Herdewijn, P., Selective
inhibitors of cyclin G associated kinase (GAK) as anti-hepatitis C
agents. J Med Chem 2015, 58, 3393-3410.
5
6
7
8
9
5. Dzamko, N.; Zhou, J.; Huang, Y.; Halliday, G. M.,
Parkinson's disease-implicated kinases in the brain; insights into
disease pathogenesis. Front Mol Neurosci 2014, 7, 57.
6. Zhang, C. X.; Engqvist-Goldstein, A. E.; Carreno, S.; Owen,
D. J.; Smythe, E.; Drubin, D. G., Multiple roles for cyclin G-
associated kinase in clathrin-mediated sorting events. Traffic
2005, 6, 1103-1113.
7. Eisenberg, E.; Greene, L. E., Multiple roles of auxilin and
hsc70 in clathrin-mediated endocytosis. Traffic 2007, 8, 640-646.
8. Chaikuad, A.; Keates, T.; Vincke, C.; Kaufholz, M.; Zenn, M.;
Zimmermann, B.; Gutierrez, C.; Zhang, R. G.; Hatzos-Skintges,
C.; Joachimiak, A.; Muyldermans, S.; Herberg, F. W.; Knapp, S.;
Muller, S., Structure of cyclin G-associated kinase (GAK) trapped
in different conformations using nanobodies. Biochem J 2014,
459, 59-69.
9. Susa, M.; Choy, E.; Liu, X.; Schwab, J.; Hornicek, F. J.;
Mankin, H.; Duan, Z., Cyclin G-associated kinase is necessary for
osteosarcoma cell proliferation and receptor trafficking. Mol
Cancer Ther 2010, 9, 3342-3350.
10. Sakurai, M. A.; Ozaki, Y.; Okuzaki, D.; Naito, Y.; Sasakura,
T.; Okamoto, A.; Tabara, H.; Inoue, T.; Hagiyama, M.; Ito, A.;
Yabuta, N.; Nojima, H., Gefitinib and luteolin cause growth
arrest of human prostate cancer PC-3 cells via inhibition of
cyclin G-associated kinase and induction of miR-630. PLoS One
2014, 9, e100124.
11. Ray, M. R.; Wafa, L. A.; Cheng, H.; Snoek, R.; Fazli, L.;
Gleave, M.; Rennie, P. S., Cyclin G-associated kinase: a novel
androgen receptor-interacting transcriptional coactivator that is
overexpressed in hormone refractory prostate cancer. Int J
Cancer 2006, 118, 1108-1119.
12. Knapp, S.; Arruda, P.; Blagg, J.; Burley, S.; Drewry, D. H.;
Edwards, A.; Fabbro, D.; Gillespie, P.; Gray, N. S.; Kuster, B.;
Lackey, K. E.; Mazzafera, P.; Tomkinson, N. C.; Willson, T. M.;
Workman, P.; Zuercher, W. J., A public-private partnership to
unlock the untargeted kinome. Nat Chem Biol 2013, 9, 3-6.
13. Arrowsmith, C. H.; Audia, J. E.; Austin, C.; Baell, J.; Bennett,
J.; Blagg, J.; Bountra, C.; Brennan, P. E.; Brown, P. J.; Bunnage, M.
E.; Buser-Doepner, C.; Campbell, R. M.; Carter, A. J.; Cohen, P.;
Copeland, R. A.; Cravatt, B.; Dahlin, J. L.; Dhanak, D.; Edwards,
A. M.; Frederiksen, M.; Frye, S. V.; Gray, N.; Grimshaw, C. E.;
Hepworth, D.; Howe, T.; Huber, K. V.; Jin, J.; Knapp, S.; Kotz, J.
D.; Kruger, R. G.; Lowe, D.; Mader, M. M.; Marsden, B.; Mueller-
Fahrnow, A.; Muller, S.; O'Hagan, R. C.; Overington, J. P.; Owen,
D. R.; Rosenberg, S. H.; Roth, B.; Ross, R.; Schapira, M.;
Schreiber, S. L.; Shoichet, B.; Sundstrom, M.; Superti-Furga, G.;
Taunton, J.; Toledo-Sherman, L.; Walpole, C.; Walters, M. A.;
Willson, T. M.; Workman, P.; Young, R. N.; Zuercher, W. J., The
promise and peril of chemical probes. Nat Chem Biol 2015, 11,
536-541.
14. Fabian, M. A.; Biggs, W. H., 3rd; Treiber, D. K.; Atteridge,
C. E.; Azimioara, M. D.; Benedetti, M. G.; Carter, T. A.; Ciceri,
P.; Edeen, P. T.; Floyd, M.; Ford, J. M.; Galvin, M.; Gerlach, J.
L.; Grotzfeld, R. M.; Herrgard, S.; Insko, D. E.; Insko, M. A.;
Lai, A. G.; Lelias, J. M.; Mehta, S. A.; Milanov, Z. V.; Velasco, A.
M.; Wodicka, L. M.; Patel, H. K.; Zarrinkar, P. P.; Lockhart, D.
J., A small molecule-kinase interaction map for clinical kinase
inhibitors. Nat Biotechnol 2005, 23, 329-336.
17. Asquith, C. R. M.; Laitinen, T.; Bennett, J. M.; Godoi, P. H.;
East, M. P.; Tizzard, G. J.; Graves, L. M.; Johnson, G. L.; Dornsife,
R. E.; Wells, C. I.; Elkins, J. M.; Willson, T. M.; Zuercher, W. J.,
Identification and optimization of 4-anilinoquinolines as
inhibitors of cyclin G associated kinase. ChemMedChem 2018, 13,
48-66.
18. Haile, P. A.; Votta, B. J.; Marquis, R. W.; Bury, M. J.;
Mehlmann, J. F.; Singhaus, R.; Charnley, A. K.; Lakdawala, A. S.;
Convery, M. A.; Lipshutz, D. B.; Desai, B. M.; Swift, B.; Capriotti,
C. A.; Berger, S. B.; Mahajan, M. K.; Reilly, M. A.; Rivera, E. J.;
Sun, H. H.; Nagilla, R.; Beal, A. M.; Finger, J. N.; Cook, M. N.;
King, B. W.; Ouellette, M. T.; Totoritis, R. D.; Pierdomenico, M.;
Negroni, A.; Stronati, L.; Cucchiara, S.; Ziółkowski, B.;
Vossenkämper, A.; MacDonald, T. T.; Gough, P. J.; Bertin, J.;
Casillas, L. N., The identification and pharmacological
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
characterization
of
6-(tert-butylsulfonyl)-N-(5-fluoro-1H-
indazol-3-yl)quinolin-4-amine (GSK583), a highly potent and
selective inhibitor of RIP2 kinase. J Med Chem 2016, 59, 4867-
4880.
19. Fedorov, O.; Niesen, F. H.; Knapp, S., Kinase inhibitor
selectivity profiling using differential scanning fluorimetry.
Methods Mol Biol 2012, 795, 109-118.
20. Vasta, J. D.; Corona, C. R.; Wilkinson, J.; Zimprich, C. A.;
Hartnett, J. R.; Ingold, M. R.; Zimmerman, K.; Machleidt, T.;
Kirkland, T. A.; Huwiler, K. G.; Ohana, R. F.; Slater, M.; Otto, P.;
Cong, M.; Wells, C. I.; Berger, B. T.; Hanke, T.; Glas, C.; Ding, K.;
Drewry, D. H.; Huber, K. V. M.; Willson, T. M.; Knapp, S.;
Müller, S.; Meisenheimer, P. L.; Fan, F.; Wood, K. V.; Robers, M.
B., Quantitative, wide-spectrum kinase profiling in live cells for
assessing the effect of cellular ATP on target engagement. Cell
Chem Biol 2018, 25, 206-214.e11.
21. Tigno-Aranjuez, J. T.; Benderitter, P.; Rombouts, F.;
Deroose, F.; Bai, X.; Mattioli, B.; Cominelli, F.; Pizarro, T. T.;
Hoflack, J.; Abbott, D. W., In vivo inhibition of RIPK2 kinase
alleviates inflammatory disease. J Biol Chem 2014, 289, 29651-
29664.
22. Casillas, L. N.; Demartino, M. P.; Haile, P.; Mehlmann, J.
F.; Ramajulu, J. M.; Singhaus, R. J. Amino-Quinolines as Kinase
Inhibitors. WO2014/043437, 2014.
23. Casillas, L. N.; Haile, P. A.; Marquis, R. W. J.; Wang, G.
Prodrugs
of
Amino
Quinazoline
Kinase
Inhibitor.
WO2014/043446, 2014.
24. Shimizu, H.; Nagamori, I.; Yabuta, N.; Nojima, H., GAK, a
regulator of clathrin-mediated membrane traffic, also controls
centrosome integrity and chromosome congression. J Cell Sci
2009, 122 (Pt 17), 3145-3152.
25. Orth, J. D.; Loewer, A.; Lahav, G.; Mitchison, T. J.,
Prolonged mitotic arrest triggers partial activation of apoptosis,
resulting in DNA damage and p53 induction. Mol Biol Cell 2012,
23, 567-576.
26. Munoz, L., Non-kinase targets of protein kinase inhibitors.
Nat Rev Drug Discov 2017, 16, 424-440.
27. Dolly, S. O.; Gurden, M. D.; Drosopoulos, K.; Clarke, P.; de
Bono, J.; Kaye, S.; Workman, P.; Linardopoulos, S. RNAi screen
reveals synthetic lethality between cyclin G-associated kinase
and FBXW7 by inducing aberrant mitoses. Br J Cancer 2017, 117,
954-964.
15. Tabara, H.; Naito, Y.; Ito, A.; Katsuma, A.; Sakurai, M. A.;
Ohno, S.; Shimizu, H.; Yabuta, N.; Nojima, H., Neonatal lethality
in knockout mice expressing the kinase-dead form of the
28. Dehm, S. M.; Tindall, D. J., Alternatively spliced androgen
receptor variants. Endocr Relat Cancer 2011, 18, R183-R196.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 10
Table of Contents Graphic
1
2
3
4
5
OMe
MeO
SGC-GAK-1
GAK DSF Tm 7.7 °C
6
7
8
9
GAK Ki = 3.1 nM
MeO
NH
N
GAK nanoBRET IC50 = 120 nM
NAK selectivity - 16000
Br
>50-fold selectivity over nearest kinase
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7
ACS Paragon Plus Environment

Page 9 of 10
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
F
8
9
NH
N
Cl
NH
N
O
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
O
O
O
N
N
OMe
OMe
N
OMe
Gefitinib, GAK KD = 13 nM
Erlotinib, GAK KD = 3.1 nM
OMe
MeO
OMe
MeO
N
MeO
S
MeO
NH
N
CF3
N
N
O
1, GAK KD = 8.9 nM
2, GAK KD = 31 nM
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 10
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

Page 11 of 10
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
ACS Paragon Plus Environment