Highly potent and isoform selective dual site binding tankyrase/Wnt signaling inhibitors that increase cellular glucose uptake and have antiproliferative activity

Article abstract of DOI:10.1021/acs.jmedchem.6b01574

Compounds 13 and 14 were evaluated against 11 PARP isoforms to reveal that both 13 and 14 were more potent and isoform selective toward inhibiting tankyrases (TNKSs) than the "standard" inhibitor 1 (XAV939)⁵, i.e., IC₅₀ = 100 pM vs TNKS2 and IC₅₀ = 6.5 μM vs PARP1 for 14. In cellular assays, 13 and 14 inhibited Wnt-signaling, enhanced insulin-stimulated glucose uptake, and inhibited the proliferation of DLD-1 colorectal adenocarcinoma cells to a greater extent than 1.

Full text of DOI:10.1021/acs.jmedchem.6b01574

Subscriber access provided by GAZI UNIV
Brief Article
Highly potent and isoform-selective dual-site-binding
tankyrase/Wnt signaling inhibitors that increase cellular
glucose uptake and have anti-proliferative activity
AMIT NATHUBHAI, Teemu Haikarainen, Jarkko Koivunen, Sudarshan
Murthy, Françoise Koumanov, Matthew D. Lloyd, Geoffrey D. Holman,
Taina Pihlajaniemi, David Tosh, Lari Lehtiö, and Michael D. Threadgill
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01574 • Publication Date (Web): 16 Dec 2016
Downloaded from http://pubs.acs.org on December 17, 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 7
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Highly potent and isoform-selective dual-site-binding tanky-
rase/Wnt signaling inhibitors that increase cellular glucose uptake
and have anti-proliferative activity.
Amit Nathubhai,*† Teemu Haikarainen,‡ Jarkko Koivunen,‡ Sudarshan Murthy,‡ Françoise Koumanov,§
Matthew D. Lloyd,† Geoffrey D. Holman,§ Taina Pihlajaniemi,‡ David Tosh,§ Lari Lehtiӧ,‡ Michael D.
Threadgill†
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
† Drug and Target Discovery, Department of Pharmacy and Pharmacology, University of Bath, Bath BA2 7AY, U. K.,
‡Faculty of Biochemistry and Molecular Medicine, Biocenter Oulu, University of Oulu, PO Box 5400, 90014 Oulu, Finland
§Department of Biology and Biochemistry, University of Bath, Bath BA2 7AY, U. K.
KEYWORDS: Tankyrase, PARP, Glucose uptake, Wnt signaling, Crystal structure, Insulin.
ABSTRACT: Compounds 13 and 14 were evaluated against eleven PARP isoforms to reveal that both 13 and 14 were more potent
and isoform-selective towards inhibiting tankyrases (TNKSs) than the “standard” inhibitor 1 (XAV939)5, i.e. IC50 = 100 pM vs.
TNKS2 and IC50 = 6.5 µM vs. PARP1 for 14. In cellular assays, 13 and 14 inhibited Wnt-signaling, enhanced insulin-stimulated
glucose uptake and inhibited the proliferation of DLD-1 colorectal adenocarcinoma cells to a greater extent than 1.
INTRODUCTION
ones such as 2,16,17 isoquinolin-1-ones18 and aryltetrahydro-
naphthyridinones, which maintain a classical binding mode in
thenicotinamide-binding site.19 Compound 3 (IWR-1)20 (Fig. 1)
is an inhibitor of the Wnt signaling cascade through inhibiting
TNKSs, binding only to the adenosine-binding site.20 The nor-
bornane of 3 forces a conformational change of Tyr1050 (TNKS2
numbering), allowing the quinoline of 3 to -stack with His1048
within this site.21 Our previous reports of novel inhibitors in-
clude 2-aryl-8-methylquinazolin-4-ones with 4’-large or elec-
tron-withdrawing substituents (e.g. 2) to provide inhibition of
TNKSs (IC50 in the low nM range) and of Wnt / -catenin sig-
naling.16,17 Some known inhibitors bind in both the nicotina-
mide-binding site and the adenosine-binding domain,22,23 alt-
hough the increases in potency vs. 1 were modest.
Tankyrases (TNKSs) are members of the poly(ADP-ribose)-
polymerase (PARP) family of seventeen enzymes that use
NAD+ as a substrate to transfer ADP-ribose units to target pro-
teins.1 There are two human isoforms, TNKS1 and TNKS2.
Target proteins include telomere repeating binding factor-1
(TRF1), 2,3 nuclear mitotic apparatus protein (NuMA)4 (essen-
tial for the resolution of chromatids during mitosis) and axin,
the limiting component of Wnt / -catenin signaling.5 Increased
Wnt signaling correlates with the overexpression of TNKSs in
several human cancers.6-9 Inhibition of TNKSs is reported to
lead to stabilization of axin and decreased nuclear -catenin-
driven proliferation of cancer cells.10 Epidemiological studies
show that patients with Type-2 diabetes are at higher risk of de-
veloping site-specific cancers11 but the precise mechanisms that
link diabetes to cancer remain unclear. The axin-TNKS-KIF3A
complex is required for insulin-stimulated translocation of
GLUT4 to the cell memembrane.12 Insulin-regulated amino-
peptidase (IRAP) is also a binding partner and target protein of
TNKSs.13 Together, IRAP and TNKSs can enhance insulin-
stimulated exocytosis of GLUT4, which could result in in-
creased uptake of glucose.13 TNKS-knockout mice display in-
creased sensitivity to insulin and reduced adiposity and pan-
PARP inhibitors have been used to investigate the role of
TNKSs in studies on the translocation of GLUT4.14 Inhibitor 1
(XAV939)5 (Fig. 1) has been extensively used as a tool to in-
hibit TNKSs and of Wnt / -catenin signaling5 but lacks the re-
quired isoform-selectivity to avoid off-target effects. Related
inhibitors of TNKSs include flavones,15 2-arylquinazolin-4-
Figure 1. Examples of previously studied inhibitors of TNKSs 1, 2
(nicotinamide-site binders) and 3 (adenosine-site binder). Inhibitor
2 is shown with locant numbers.
Here, we report the rational design and evaluation of advanced
inhibitors (13, 14) which maintain the 8-methylquinazolin-4-
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 7
1
2
3
4
5
6
7
8
one core and carry extensions at the 2-position to induce con-
clisation and hydrolysis in aqueous base led to the quinazoli-
formational change and binding to the adenosine-binding site,
leading to a step-change in potency and isoform-selectivity.
These highly potent and isoform-selective inhibitors have po-
tential as excellent molecular probes applicable to research on
diabetes and cancer.
nonepropanoic acids 11 and 12. Careful optimization of the
coupling conditions was required to join 11 / 12 with 6; simul-
taneous addition of the activating agent (CDI) and quinazoli-
nones 11 / 12 to 6 was essential to provide the candidate dual-
site-binding inhibitors 13 and 14 in modest yields.
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
RESULTS AND DISCUSSION
Modeling. Structural alignment of the co-crystal structures of 2
(PDB 4UFU) and 3 (PDB 3UA9) with TNKS217,21 provided in-
itial insights towards the design of 13 and 14. The quinazoli-
none of 2 binds in the nicotinamide-binding site, making the
expected H-bonds with Ser1068 and Gly1032 and -stack with
Tyr1071 (TNKS2 numbering). The quinoline of 3 is located in
the adenosine-binding site, making a -stack with His1048
(TNKS2 numbering). Features binding at each site were linked
to create chimeric compounds 13 and 14, which combine im-
portant H-bonds and stacking interactions at both binding sites.
Modeling of 13 and 14 into the active site of TNKS1 (PDB
4I9I)23 predicted that the designed compounds could bind to the
pockets in the intended way and that the length and nature of
the linker were appropriate (Fig. 2). The quinazolin-4-one core
could bind in the nicotinamide-binding site to make the classi-
cal H-bond and -stacking interactions. Compounds 1 and 2
contain a 2-aryl group, which is shown to occupya hydrophobic
cavity. Modeling of 13 into TNKS1 suggests that the chosen
linker could allow Tyr1203 to move towards the nicotinamide-
binding site and decrease the volume of the hydrophobic cavity.
However, this shrinkage of the pocket could still allow the pro-
panamide linker to thread through. The modeling study reveals
that the linker could interact further with the protein, in that the
C=O of the propanamide linker is appositely located to H-bond
with the backbone NH of Tyr1213. Moreover, the central benzene
ring and the quinoline of 13 are predicted to cause movement
of His1201 of the adenosine-binding site to place its imidazole
appropriately for -stacking with the quinoline. We have previ-
ously reported that an 8-Me group enhances inhibitory activity
in the simple 2-arylquinazolinones16,17 and related 3-aryli-
soquinolinones,18 we also incorporated the 8-Me in 14, to test if
it would still make a contribution in these advanced dual-site-
binding compounds (Fig. 2).
Scheme 1. Synthesis of 13 and 14. Reagents and conditions: i)
-
nitrobenzoyl chloride, pyridine, THF, 16 h, Ar; ii) +NH4 HCO2
,
10% Pd/C, DMF/MeOH (2:1),
2
h, Ar; iii)
EtO2CCH2CH2COCl, pyridine, THF, 16 h; iv) aq. NaOH (0.5
M), 60°C; v) 6, DMF, Pri2NEt, carbonyldiimidazole, 72 h, Ar.
Biochemical evaluation. The target compounds 13 and 14
were evaluated in vitro for inhibition of TNKS1 and TNKS2
and counterscreened against eleven PARP isoforms, including
the major isoform PARP1 (Table 1). Known inhibitors 1 and 3
were also examined as standards against PARP1, PARP2,
TNKS1 and TNKS2. The nicotinamide-site binder 1 showed
IC50 values in the low nanomolar range for inhibition of TNKS1
and TNKS2, which were similar to those reported by Huang et
al.5 However, although 1 showed reasonable selectivity for
TNKS2 vs. PARP1 (131-fold), other comparisons called into
question the use of this agent as a selective inhibitor (TNKS1
vs. PARP1 25-fold; TNKS1 vs. PARP2 4.1-fold). Compound 3,
which binds in the adenosine-binding site, had moderate activ-
ity against TNKS1 and TNKS2 (IC50 = 343 nM and 31 nM, re-
spectively) and did not inhibit PARP1 or PARP2 up to 10 M.
These data are again consistent with earlier evaluations of the
selectivity of this agent. Huang et al.5 reported IC50 (TNKS2) =
56 nM and selectivity vs. PARP1 and PARP2 > 300-fold, while
Narwal et al. confirmed the selectivity of 3 with IC50 = ca. 100
µM (PARP1) and IC50 = ca. 35 µM (PARP2).21 Thus binding at
the adenosine site has potential for greater selectivity for
TNKSs vs. other PARPs, although this may not give high po-
tency as the sole binding site.
Intermediate 6, containing the putative adenosine-site-binding
benzamidoquinoline, was evaluated for ability to bind in the ab-
sence of the anchoring quinazolin-4-one but it failed to inhibit
any of the isoforms. This shows that the norbornane of 3 is es-
sential, interacting with Tyr1203 to modify the geometry of the
adenosine-binding site to accept the benzamidoquinoline.
Figure 2. Model of 13 docked into the structure of TNKS1 (PDB
code 4I9I).23 The protein is shown in pink and 13 in cyan. Key H-
bonds shown as grey dashed lines.
Chemical Synthesis. The synthesis of target compounds (Sch.
1) began by acylation of 8-aminoquinoline 4 with 4-nitrobenz-
oyl chloride to give 5, followed by transfer hydrogenation to
provide amine 6. Anthranilamides 7 and 824 were acylated with
ethyl 4-chloro-4-oxobutanoate to provide 9 and 10. One-pot cy-
The candidate dual-site inhibitors 13 and 14 were shown to be
extremely potent and isoform-selective inhibitors of TNKSs, in-
hibiting TNKS2 in the pM range and thus they are both
2
ACS Paragon Plus Environment

Page 3 of 7
Journal of Medicinal Chemistry
1
2
3
Table 1. IC50 (pIC50 ± standard error) values for inhibition of TNKS1, TNKS2 and PARP isoforms in biochemical assays and for
4
inhibition of Wnt signaling in a cellular assay.
5
6
7
8
Compound
PARP1
1
3
6
11
6700 nM
(5.18 ± 0.17)
125 nM
12
1000 nM
(5.18 ± 0.17)
145 nM
13
8230 nM
(5.08 ± 0.12)
6900 nM
(5.16 ± 0.12)
40000 nM
(4.40 ± 0.14)
69000 nM
(4.16 ± 0.10)
9.1 nM
14
6500 nM
a
854 nM
(6.07 ± 0.13)
141 nM
(6.85 ± 0.08)
>100000 nM a >100000 nM
(5.19 ± 0.06)
11600 nM
(4.94 ± 0.05)
45000 nM
(4.34 ± 0.10)
65000 nM
(4.19 ± 0.07)
5.1 nM
>100000 nM a
b
PARP2
PARP3
PARP4
TNKS1
TNKS2
9
(6.90 ± 0.04)
(6.84 ± 0.12)
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
a
34 nM
(7.47 ± 0.23)
6.5 nM
343 nM
(6.46 ± 0.09)
31 nM
>100000 nM
>100000 nM
>100000 nM a >100000 nM a
>100000 nM a >100000 nM a
(8.04 ± 0.05)
0.20 nM
(9.69 ± 0.17)
(8.29 ± 0.27)
0.10 nM
(10.0 ± 0.05)
(8.19 ± 0.09)
(7.51 ± 0.10)
>100000 nM a >100000 nM a
>100000 nM a >100000 nM a
>100000 nM a >100000 nM a
>100000 nM a >100000 nM a
>100000 nM a >100000 nM a
PARP10
PARP12
PARP14
PARP15
PARP16
Wnt
220nM
(6.66 ± 0.10)
136 nMc
29 nM
(7.54 ± 0.05)
37 nM
(7.43 ± 0.06)
a Limited by solubility. b Empty cells indicate not determined. c Determined previously.23
considerably more potent than the “standard” inhibitor1 and the
adenosine-site-binder 3. This increase in potency of inhibition
of the TNKSs for 13 and 14 was not accompanied by an in-
crease in activity against other members of the PARP family
(Table 1). In particular, 13 and 14 showed very weak activity
against PARP1 and PARP2 (IC50 >> 6 M), leading to exquisite
isoform-selectivity. Inhibitor 13 is 4.1  104-fold selective for
TNKS2 vs. PARP1 and 3.5  104-fold selective for TNKS2 vs.
PARP2. Similarly, 14 is 6.5  104-fold selective for TNKS2 vs.
PARP1 and 11.6  104-fold selective for TNKS2 vs. PARP2.
The somewhat weaker potency of these agents against TNKS1
corresponds to selectivities for TNKS1 vs. PARP1 and PARP2
in the range 1-2.3  103-fold. Thus exploiting dual-site binding
with an appropriate linker makes a step-change in potency and
provides isoform-selectivity with this potency. Compound 14 is
approximately twice as potent as 13 against both isoforms, con-
firming the moderate advantage of the 8-Me on the quinazoli-
none core.16,18,17 Intermediates 11 and 12, carrying only a pro-
panoic acid at the quinazolinone 2-position, were evaluated to
explore the contribution of the quinazolinone-CH2CH2CO unit
towards potency and selectivity; both showed no inhibition of
TNKSs. We have previously shown that a 2-aryl group is re-
quired for potent inhibition of TNKSs; this requirement is
thereby reinforced for inhibitors that bind only as mimics of
nicotinamide.16-19 However, 11 and 12 do inhibit PARP1 and
PARP2. Previously, we showed that polar groups at the 4’-po-
sition of 2-arylquinazolin-4-ones decreased selectivity for
TNKSs, in that inhibition of PARP1 was enhanced;16-18 there-
fore, the 2-propanoic acid of 11 and 12 is well suited to interact
with the corresponding region of protein of PARP1 that con-
tains hydrophilic residues. Furthermore, 11 and 12 displayed
moderately potent and selective inhibition of PARP2 (54- and
6.9-fold, respectively, vs. PARP1), results which place 11
amongst the most PARP2-selective agents known.25 Therefore,
11 and 12 provide a new scaffold towards the development of
novel PARP2-selective inhibitors.
form-selectivity of 13 and 14, crystal structures of the com-
pounds in complex with TNKS2 catalytic domain were solved
(Supplementary information). The quinazolinone binds to the
nicotinamide pocket as designed and forms interactions with the
protein similar to those of 2 (PDB code 4UFU).17 The stacking
interaction with Tyr1071 and the H-bonds with Gly1032 and Ser1068
are preserved (Fig. 3). Binding of 3 causes opening of the ac-
tive-site loop and moves His1048 away from the binding
pocket.21 Similar changes are observed in the crystal structures
of the complexes of 13 and 14 and the quinoline binds to form
an induced adenosine-binding pocket. Importantly, the H-bonds
to Tyr1060 and to Asp1045, observed in adenosine-site inhibitors,26
are both present in the TNKS2 structures containing 13 and 14.
The central phenyl ring adopts slightly different conformations
in the experimental crystal structures of 13 and 14, indicating
flexibility in the hydrophobic surface at this site. A recent dual-
site binding inhibitor binds to the catalytic domain in a very
similar manner (IC50 = 2 nM vs. TNKS2)23 but does not create
as efficient π-stacking interaction with His1201 as does the elec-
tron-poor bicycle of 13 and 14. This may explain why com-
pounds 13 and 14 exhibit such high inhibitory activity towards
TNKSs. Unlike PARP1 and PARP2, TNKSs do not contain an
-helical regulatory domain. This domain packs against the
NAD+-binding groove and the interactions at the adenosine site
contribute to the potent selectivity of the adenosine-site binders.
The extreme isoform-selectivity of 13 and 14 can be rational-
ized, similarly to 3, to be mainly due to the quinoline in the
adenosine-binding pocket and unfavorable interactions with the
charged residues present in PARP-1 and PARP-2.17,21
X-ray crystallography. To rationalize the potency and iso-
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 7
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 3. Crystal structure of A) TNKS2-13 (PDB code 5FPF) and
B) TNKS2-14 (PDB code 5FPG). Hydrogen bonds are shown in
dashed lines and the electron density (2mFo-DFc) is contoured at 1
Å around the ligand.
Inhibition of Wnt signaling in TCF/LEF reporter-HEK293
cells. The intracellular activity and potency of 13 and 14 was
evaluated in a functional assay using a TCF/LEF Reporter-
HEK293 cell line (BPS Bioscience, catalog #60501).27 Upon
extracellular Wnt signals, -catenin is stabilized and associates
with TCF/LEF transcription factors, activating transcription.5,27
TNKSs control the stability of the -catenin destruction com-
plex and inhibition is expected to stabilize the destruction com-
plex and thus lower the levels of β-catenin.5 The TCF/LEF Re-
porter-HEK293 cell line is used to measure this interference
with this pathway. Both 13 (IC50 = 29 nM) and 14 (IC50 = 37
nM) were shown to be extremely effective and potent cellular
inhibitors of Wnt / -catenin signaling, confirming their uptake
into cells and their activity therein (Table 1). Thus 13 and 14
are ca. 10 more potent in cells than the standard inhibitor 1,
reflecting the greater potency of the dual-site inhibitors in vitro.
Cell viability was monitored with a light microscope; 13 and 14
were not cytotoxic in this assay.
Figure 4. Upper: Exemplary image of DLD-1 colony forming assay
(1000 cells/well) with 1, 3, 13 and 14 (0 M (control with 1%
DMSO v/v only), 1.0 M and 100 nM). Lower: Histogram of data
from assay. 1, 3, 13 and 14 (n = 3).
Insulin-stimulated glucose uptake. The axin-TNKS-KIF3A
complex is stabilized through inhibition of TNKS.12 Ablation of
expression of TNKSs has been reported to upregulate GLUT4
at the post-transcriptional level, potentially increasing uptake of
glucose into adipocytes.14 To address this pharmacologically,
13 and 14 were examined for their ability to enhance insulin-
stimulated uptake of glucose. 3T3-L1 Adipocytes were treated
with 13 and 14 in the presence of insulin (100 nM).29 Compound
1 was used for comparison as the current “standard” TNKS in-
hibitor. The insulin-stimulated uptake of radiolabeled 2-deoxy-
D-glucose into the cells was measured. The weaker TNKS in-
hibitor 1 had no effect on insulin-stimulated glucose uptake in
comparison to the insulin-only control. The slight decrease in
non-stimulated glucose uptake in presence of 1 (1.0 M) was
not significantly different from control. Non-stimulated glucose
uptake in presence of 13 and 14 (1.0 M) was not significantly
different from control. In comparison to insulin only and with
Anti-proliferative activity towards DLD-1 human colon car-
cinoma cells. Aberrant Wnt signaling is found in > 90% of col-
orectal cancers, owing to mutation in the APC protein which is
a component of the β-catenin destruction complex.6 Compound
1 has been reported to have anti-proliferative activity against
DLD-1 human colon cancer cells but only under serum-de-
prived conditions.5,28 Using a colony-forming assay, 13 and 14
were shown to inhibit formation of colonies of DLD-1 cells
even under normal-serum conditions (Fig. 4) at 100 nM and 1.0
M, conditions under which 1 was inactive. This observation
again reflects the greater potency of 13 and 14 in vitro and in
cells.
4
ACS Paragon Plus Environment

Page 5 of 7
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1, both 13 and 14 significantly further increased insulin-stimu-
lated glucose uptake. Glucose uptake increased by ca. 40% and
20% using 100 nM and 1.0 µM of 13, respectively. Insulin-
stimulated uptake of glucose increased by ca. 20% by 14 at 100
nM and 1.0 µM in comparison to insulin treatment only (Fig.
5). This confirms potent intracellular activity of 13 and 14.
TNKS-KIF3A complex.12 In comparative studies in vitro, the
widely-used TNKS inhibitor 1 has been shown to be markedly
less potent than 13 and 14 and to lack the exquisite isoform-
selectivity of these new agents. The improved potency and iso-
form-selectivity of 13 and 14 is manifest in three comparative
cellular functional assays, where they show useful activity in
situations where 1 lacks potency or activity.
Compounds 13 and 14 are now available as potent and isoform-
selective TNKS inhibitors which inhibit the Wnt response phar-
macologically in cells. Theywill have applications as molecular
tools in studies on Wnt and related systems. These results also
indicate that inhibition of TNKSs can increase sensitivity of
cells to insulin, with potential application in Type-2 diabetes, as
high doses of insulin can cause dysregulation of various signal-
ing cascades (PI3K / Akt / mTOR).30 Antiproliferative activity
has also been demonstrated in human colon cancer cells under
normal serum conditions, supporting TNKS as a therapeutic tar-
get in the many cancers with aberrant Wnt signaling. A new
structural scaffold (11, 12) for selective inhibitors of PARP2
has also been identified.
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
160
*
140
*
*
120
100
80
60
40
20
0
EXPERIMENTAL SECTION
Chemistry
The purity of target compounds were >95% as determined by
1H and 13C NMR, high resolution mass spectrometry using elec-
trospray ionization and HPLC analysis at two different wave-
lengths (see supplementary information for general experi-
mental, spectra of all compounds and synthetic procedures of
intermediate compounds).
Synthesis of target compounds 13 and 14.
2-(3-Oxo-3-(4-((quinolin-8-yl)aminocarbonyl)phenyla-
mino)propyl)quinazolin-4-one (13). Compound 6 (548 mg,
2.08 mmol) in dry DMF (50 mL) was treated with Pri2NEt (2.96
g, 22.9 mmol), then carbonyldiimidazole (371 mg, 2.29 mmol),
followed by addition of 11 (500 mg, 2.29 mmol). The mixture
was stirred for 72 h under Ar. The solvent was evaporated and
the residue was dissolved in EtOAc/MeOH (1:2, 70 mL). The
organic solution was washed with water (3  30 mL) and brine
(3  30 mL), then dried. Evaporation and chromatography (1:9
EtOAc / CH2Cl2 → 1:9 MeOH / EtOAc) gave a mixture con-
taining 13. Further chromatography (EtOAc) gave 13 (74 mg,
8%) as a pale purple solid: mp 265-268°C 1H NMR ((CD3)2SO)
2.95-2.98 (4 H, m), 7.47 (1 H, t, J = 7.0 Hz), 7.63 (1 H, d, J =
8.0 Hz), 7.64-7.69 (2 H, m), 7.72-7.83 (2 H, m), 7.82 (2 H, d, J
= 8.5 Hz), 8.00 (2 H, d, J = 8.5 Hz), 8.08 (1 H, d, J = 8.0 Hz),
8.46 (1 H, dd, J = 8.0, 1.5 Hz), 8.72 (1 H, d, J = 7.5 Hz), 8.97
(1 H, dd, J = 4.0, 1.5 Hz), 10.45 (1 H, s), 10.59 (1 H, s), 12.27
(1 H, s); 13C NMR ((CD3)2SO) 29.09, 32.34, 116.44, 118.70,
120.93, 122.13, 122.38, 125.77, 126.04, 126.76, 127.11,
127.86, 128.12, 128.44, 134.16, 134.35, 136.81, 138.26,
142.76, 148.74, 149.18, 156.52, 161.64, 163.96, 170.73; MS
m/z 464.1722 [M + H]+ (C27H22N5O3 requires 464.1723).
Figure 5. Effect of 1, 13 and 14 on % insulin-stimulated glucose
uptake in 3T3-L1 adipocytes. Basal levels (B) with or without com-
pound contain no insulin (I). I (with and without compound) ad-
ministered at final concentration 100 nM. Results are mean and
SEM relative to I only (n = 3 for 1 and n = 5 for 13 and 14). I + 100
nM 13 vs. I p = 0.02; I + 1.0 M 13 vs. I p = 0.02; I + 100 nM 14
vs. I p = 0.07; I + 1.0 M 14 vs. I p = 0.01; B vs. B + 13 p = 0.98;
B vs. B + 14 p = 0.94; B vs. B + 1 p = 0.21.
CONCLUSION
In this paper, we disclose the design, synthesis and evaluation
of two new highly potent and isoform-selective TNKS inhibi-
tors, 13 and 14. Crystal structures of these compounds bound
into the catalytic domain of TNKS2 confirmed that, as de-
signed, thequinazolin-4-one moiety occupied thenicotinamide-
binding site, setting up the linker so that the quinoline moiety
interacts with the adenosine-binding region. This dual-site de-
sign has led to the most potent and isoform-selective inhibitors
reported to date, with IC50 = 100 pM for inhibition of TNKS2
by 14 and 1.2  105-fold selectivity for inhibition of TNKS2 vs.
the major PARP isoform, PARP1. Cellular uptake of these
agents was demonstrated by their potent inhibition of Wnt / -
catenin signaling in the low nM range. Significant anti-prolifer-
ative activity was demonstrated in DLD-1 human colon carci-
noma cells at 100 nM under normal serum conditions. For the
first time, truly potent and selective TNKS inhibitors are shown
to increase insulin-stimulated glucose uptake, whereas previous
studies have used non-selective pan-PARP inhibitors;13 this
provides further evidence of a role of TNKSs in insulin-stimu-
lated glucose transport and further studies will elucidate the
mechanisms of this activation, presumably through the axin-
8-Methyl-2-(3-oxo-3-(4-((quinolin-8-yl)aminocarbonyl)phe-
nylamino)propyl)quinazolin-4-one (14). Compound 6 (514
mg, 1.95 mmol) in dry DMF (50 mL) was treated with Pri2NEt
(2.78 g, 21.5 mmol), then carbonyldiimidazole (348 mg, 2.15
mmol), followed by addition of 12 (500 mg, 2.15 mmol). The
mixture was stirred for 72 h under Ar. The solvent was evapo-
rated and the residue was dissolved in EtOAc/MeOH (1:2, 70
mL). The solution was washed with water (3  30 mL) and brine
(3  30 mL), then dried. Evaporation and chromatography (1:9
EtOAc / CH2Cl2 → EtOAc) gave 14 (119 mg, 21%) as a beige
solid: mp 280-282°C; 1H NMR ((CD3)2SO)  2.44 (3H, s), 2.93
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 7
1
2
3
4
5
6
7
8
9
ID23-1 and ID29. We thank Dr. Silvia Muñoz-Descalzo (Univer-
sity of Bath) and Dr. Penelope C. Hayward (University of Cam-
bridge) for helpful discussions.
(2 H, t, J = 6.0 Hz), 3.01 (2 H, t, J = 7.0 Hz), 7.31 (1 H, t, J =
7.5 Hz), 7.58 (1 H, m), 7.63-7.69 (2 H, m), 7.70 (1 H, dd, J =
8.5, 1.0 Hz), 7.83 (2 H, d, J = 8.5 Hz), 7.91 (1 H, dd, J = 8.0,
1.0 Hz), 8.00 (2 H, d, J = 8.5 Hz), 8.46 (1 H, dd, J = 8.5, 2.0
Hz), 8.72 (1 H, dd, J = 7.5, 1.5 Hz), 8.97 (1 H, dd, J = 4.5, 2.0
Hz), 10.76 (1 H, s), 10.59 (1 H, s), 12.25 (1 H, s); 13C NMR
((CD3)2SO) 17.06, 28.99, 32.08, 116.45, 118.67, 120.76,
122.11, 122.38, 123.36, 125.44, 127.11, 127.86, 128.09,
128.34, 134.17, 134.59, 134.84, 136.81, 138.26, 142.91,
147.07, 149.17, 155.12, 161.94, 163.98, 170.98; MS m/z
478.1876 [M + H]+ (C28H24N5O3 requires 478.1879).
Funding Sources
The research was funded by Worldwide Cancer Research (Grant
No. 13-1021 to AN, MDL, DT and MDT), MRC UK (Grant No.
MR/J003417/1) to FK and GDH), Biocenter Oulu, Academy of
Finland (287063 and 294085 to LL and TH) and the Center of Ex-
cellence Grant 2012-2017 of the Academy of Finland (284605 to
TP).
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
Counter-screening against PARP3, PARP4, PARP10,
PARP12, PARP14, PARP15 and PARP16. Compounds 13
and 14 were counter-screened for inhibition of these isoforms
ABBREVIATIONS
NAD+, Nicotinamide Adenosine Dinucleotide; ADP, Adenosine
diphosphate; GLUT4, Glucose Transporter Type-4; CDI, N,N’-
Carbonyldiimidazole; IC50, half-maximal inhibitory concentration;
PDB, Protein Data Bank; TCF, T-cell factor; LEF, Lymphoid En-
hancer-binding Factor; DMEM, Dulbecco’s Modified Eagle Me-
dium; BCA, Bicinchoninic acid.
32
using methods described previously.31, The new inhibitors
show no structural alerts for PAINS and are colorless and non-
fluorescent.
Insulin-stimulated glucose uptake. 3T3-L1 fibroblasts (from
the American Type Culture Collection), were cultured in
DMEM and differentiated to adipocytes by treatment with in-
sulin, dexamethasone and isobutylmethylxanthine, as described
previously.29 On the day of the experiment, 10-12 d post-differ-
entiation, the cells were incubated with serum-free DMEM for
2 h at 37C. Cells in the treatment group were treated with in-
creasing concentrations of 13 or 14 for 1 h. At the end of the
incubation period, the cells were washed thrice with Krebs-
Ringer-HEPES (KRH) buffer (140 mM NaCl, 4.7 mM KCl, 2.5
mM CaCl2, 1.25 mM MgSO4, 2.5 mM NaH2PO4, 10 mM
HEPES, (pH 7.4)) and incubated for 30 min with 13 or 14 and
in either the absence or presence of insulin (100 nM) at 37C.
After the 30 min incubation period, 2-deoxy-D-[2,6-3H]glucose
(final concentration 50 M, 0.1 Ci well-1) was added for 5 min
and the cells were washed four times with ice-cold KRH buffer.
Nonspecific uptake of 2-deoxy-D-glucose was measured in the
presence of 10 M cytochalasin B. The cells were lysed in aq.
NaOH (0.1 M) and radioactivity was counted in a TriCarb Pack-
ard scintillation counter (Perkin-Elmer). The concentrations of
proteins were measured using BCA protein assay kit (Thermo
Fisher Scientific). Results were calculated as nmol of 2-deoxy-
D-glucose min-1 (mg protein)-1 and are expressed as % of insu-
lin-only control. Statistical analysis; results were analyzed us-
ing two-tailed paired t tests.
ASSOCIATED CONTENT
1
Supporting Information. Synthetic details, H NMR, 13C NMR,
HRMS, HPLC, TNKS1 enzyme assay method, TNKS1 IC50
graphs, TNKS2 enzyme assay method, TNKS2 IC50 graphs,
PARP1 enzyme assay method, PARP1 IC50 graphs, PARP2 en-
zyme assay method, PARP2 IC50 graphs, Wnt signaling inhibition
cellular assay method, Wnt signaling inhibition IC50 graphs, X-ray
crystallography refinement data, colony-forming cellular assays
experimental and Molecular Formula Strings. This material is
available free of charge via the Internet at http://pubs.acs.org.
REFERENCES
Ferris, D. V. Evolution of poly(ADP-ribose) polymerase-1 (PARP-
1) inhibitors. J. Med. Chem. 2010, 53, 4561-4584.
Mi Kyung, K.; Smith, S. Persistent telomere cohesion triggers a
prolonged anaphase. Mol. Biol. Cell 2014. 25, 30-40.
Smith, S.; Giriat, I.; Schmitt, A.; de Lange, T. Tankyrase, a
poly(ADP-ribose) polymerase at the human telomeres. Science,
1998, 282, 1484-1487.
Chang, W.; Dynek, J. N.; Smith, S. NuMA is a major acceptor of
poly(ADP-ribosyl)ation by tankyrase 1 in mitosis. Biochem. J.
2005, 391, 177-184.
Huang, S.-M. A.; Mishina, Y. M.; Liu, S.; Cheung, A.; Stegmeier,
F.; Michaud, G. A.; Charlat, O.; Wiellette, E.; Zhang, Y.; Wiessner,
S.; Hild, M.; Shi, X.; Wilson, C. J.; Mickanin, C.; Myer, V.; Fazal,
A.; Tomlinson, R.; Serluca, F.; Shao, W.; Cheng, H.; Shultz, M.;
Rau, C.; Schirle, M.; Schlegl, J.; Ghidelli, S.; Fawell, S.; Lu, C.;
Curtis, D.; Kirschner, M. W.; Lengauer, C.; Finan, P. M.; Tallarico,
J. A.; Bouwmeester, T.; Porter, J. A.; Bauer, A.; Cong, F. Tanky-
rase inhibition stabilizes axin and antagonizes Wnt signalling. Na-
ture, 2009, 461, 614-620.
White, B. D.; Chien, A. J.; Dawson, D. W. Dysregulation of Wnt/-
catenin signaling in gastrointestinal cancers. Gastroenterology,
2012, 142, 219-232.
Katoh, M.; Kirikoshi, H.; Terasaki, H.; Shiokawa, K. WNT2B2
mRNA, up-regulated in primary gastric cancer, is a positive regu-
lator of the WNT--catenin-TCF signaling pathway. Biochem. Bi-
ophys. Res. Commun. 2001, 289, 1093-1098.
Miyoshi, Y.; Iwao, K.; Nagasawa, Y.; Aihara, T.; Sasaki, Y.;
Imaoka, S.; Murata, M.; Shimano, T.; Nakamura, Y. Activation of
the -catenin gene in primary hepatocellular carcinomas by so-
matic alterations involving exon 3. Cancer Res. 1998, 58, 2524-
2527.
Katoh M. Expression and regulation of WNT1 in human cancer:
Up-regulation of WNT1 by -estradiol in MCF-7 cells. Int. J. On-
col. 2003, 22, 209-212.
AUTHOR INFORMATION
Corresponding Author
* Phone +44 (0)1225 383379. E-mail: A.Nathubhai@bath.ac.uk.
Author Contributions
The manuscript was written through contributions of all au-
thors. The authors declare no competing financial interest.
Accession codes
Coordinates and structure factors are deposited at the Protein
Data Bank with codes 5FPF and 5FPG. Authors will release the
atomic coordinates and experimental data upon article publica-
tion.
ACKNOWLEDGMENT
We acknowledge the support from the European Synchrotron Ra-
diation Facility (ESRF, Grenoble, France). We are grateful to Local
Contracts at ESRF for providing assistance in using beamlines
McGonigle, S.; Chen, Z-H.; Wu, J-Y.; Chang, P.; Kolber-Simonds,
D.; Ackermann, K.; Twine, N. C.; Shie, J. L.; Miu, J. Z. T.; Huang,
K. C.; Moniz, G. A.; Nomoto, K. E7499: A dual inhibitor of
6
ACS Paragon Plus Environment

Page 7 of 7
Journal of Medicinal Chemistry
1
2
3
4
5
6
PARP1/2 and tankyrase1/2 inhibits growth of DNA repair deficient
tumors and antagonizes Wnt signalling. Oncotarget, 2015, 6,
41307-41323.
Shikata, K.; Ninomiya, T.; Kiyohara, Y.; Diabetes mellitus and
cancer risk: review of the epidemiological evidence. Cancer Sci.
2013. 104, 9-14.
Majumdar, D.; Ochala, E.; Palermo, M. G.; Pham, L.; Pu, M.;
Smith, T.; Stams, T.; Tomlinson, R. C.; Touré, B. B.; Visser, M.;
Wang, R. M.; Waters, N. J.; Shao, W. Identification of NVP-
TNKS656: The use of structure–efficiency relationships to gener-
ate a highly potent, selective, and orally active tankyrase inhibitor.
J. Med. Chem. 2013, 56, 6495-6511.
7
Guo, H.-L.; Zhang, C.; Liu, Q.; Li, Q. X; Lian, G.; Wu, D.; Li, X.;
Zhang, W.; Shen, Y.; Ye, Z.; Lin, S.-Y.; Lin, S.-C. The
Axin/TNKS complex interacts with KIF3A and is required for in-
sulin-stimulated GLUT4 translocation. Cell. Res. 2012, 22, 1246-
1257.
Yeh, T-Y. J.; Sbodio, I. J.; Tsun, Z-Y.; Lou, B.; Chi, N-W. Insulin-
stimulated exocytosis of GLUT4 is enhanced by IRAP and its part-
ner tankyrase. Biochem. J. 2007, 402, 279-290.
Yeh, T-Y. J.; Beiswenger, K. K.; Li, P.; Bolin, K. E.; Lee, R. M.;
Tsao, T-S.; Murphy, A. N.; Henever, A. L.; Chi, N-W. Hyperme-
tabolism, hyperphagia, and reduced adiposity in tankyrase-defi-
cient mice. Diabetes, 2009, 58, 2476-2485.
Narwal, M.; Haikarainen, T.; Fallarero, A.; Vuorela, P. M.; Lehtiö,
L. Screening and structural analysis of flavones inhibiting tanky-
rases. J. Med. Chem. 2013, 56, 3507-3517.
Nathubhai, A.; Wood, P. J.; Lloyd, M. D.; Thompson, A. S.;
Threadgill, M. D. Design and discovery of 2-arylquinazolin-4-ones
as potent and selective inhibitors of tankyrases. ACS Med. Chem.
Lett. 2013, 4, 1173-1177.
Paine, H. A.; Nathubhai, A.; Woon, E. C. Y.; Sunderland, P. T.;
Wood, P. J.; Mahon, M. F.; Lloyd, M. D. Thompson, A. S.;
Haikarainen, T.; Narwal, M.; Lehtiö, L.; Threadgill, M. D. Explo-
ration of the nicotinamide-binding site of the tankyrases, identify-
ing 3-arylisoquinolin-1-ones as potent and selective inhibitors in
vitro. Bioorg. Med. Chem. 2015, 23, 5891-5908.
Kumpan, K.; Nathubhai, A.; Zhang C.; Wood, P. J.; Mahon, M. F.;
Lloyd, M. D.; Thompson, A. S.; Haikarainen, T.; Lehtiö, L.;
Threadgill, M. D. Structure-based design, synthesis and evaluation
in vitro of arylnapthyridinones, arylpyridopyrimidinones and their
tetrahydro derivatives as inhibitors of the tankyrases. Bioorg. Med.
Chem. 2015, 23, 3013-3032.
Gunaydin, H.; Gu, Y.; Huang, X. Novel binding mode of a potent
and selective tankyrase inhibitor. Plos One, 2012, 7, e33740.
Narwal, M.; Venkannagari, H.; Lehtiö, L. Structural basis of selec-
tive inhibition of human tankyrases. J. Med. Chem. 2012, 55, 1360-
1367.
Bregman. H.; Gunyadin. H.; Gu. Y.; Schneider. S.; Wilson. C.; Di-
Mauro. E. F.; Huang. X. Discovery of a class of novel tankyrase
inhibitors that bind to both the nicotinamide pocket and the induced
pocket. J. Med. Chem. 2013, 56, 1341-1345.
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
Nathubhai, A.; Patterson, R.; Woodman, T. J.; Sharp, H. E. C.;
Chui, M. T. Y.; Chung, H. H. K.; Lau, S. W. S.; Zheng, J.; Lloyd,
M. D.; Thompson, A. S.; Threadgill, M. D. N3-Alkylation during
formation of quinazolin-4-ones from condensation of anthranila-
mides and orthoamides. Org. Biomol. Chem. 2011, 9, 6089-6099.
Sunderland, P. T.; Woon, E. C. Y.; Dhami, A.; Bergin, A. B.; Ma-
hon, M. F.; Wood, P. J.; Jones, L. A.; Tully, S. R.; Lloyd, M. D.;
Thompson, A. S.; Javaid, H.; Martin, N. M. B.; Threadgill, M. D.
5-Benzamidoisoquinolin-1-ones and 5-(-carboxyalkyl)isoquinol-
in-1-ones as isoform-selective inhibitors of poly(ADP-
ribose)polymerase-2 (PARP-2). J. Med. Chem. 2011, 54, 2049-
2059.
Haikarainen, T.; Waaler, J.; Ignatev, A.; Nkizinkiko, Y.; Venkan-
nagari, H.; Obaji, E.; Krauss, S.; Lehtiö, L. Development and struc-
tural analysis of adenosine site binding tankyrase inhbitors. Bioorg.
Med. Chem. Lett. 2016, 26, 328-333.
Haikarainen, T.; Koivunen, J.; Narwal. M.; Venkannagari, H.;
Obaji, E.; Joensuu, P.; Pihlajaniemi, T.; Lehtiö, L. Para-substituted
2-phenyl-3,4-dihydroquinazolin-4-ones as potent and selective
tankyrase inhibitors. ChemMedChem, 2013, 8, 1978-1985.
Bao, R; Christova, T.; Song, S.; Angers, S.; Yan, X.; Attisano, L.
Inhibition of tankyrases induces axin stabilization and block Wnt
signalling in breast cancer cells. PloS One, 2012, 7, e48670.
Frost, S. C.; Lane, M. D. Evidence for the involvement of vicinal
sulfhydryl-groups in insulin-activated hexose-transport by 3T3-L1
adipocytes. J. Biol. Chem. 1985, 260, 2646-2652.
Othman, E. M.; Hintzsche, H.; Stopper, H. Signalling steps in the
induction of genomic damage by insulin in colon and kidney cells.
Free. Rad. Biol. Med. 2014, 68, 247-257.
Venkannagari, H.; Verheugd, P.; Koivunen, J.; Pihlajaniemi, T.;
Lüscher, B.; Lehtiö, L. Small-molecule chemical probe rescues
cells from mono-ADP-ribosyltransferase ARTD10/PARP10-
induced apoptosis and sensitizes cancer cells to DNA damage. Cell
Chem. Biol. 2016, 23, 1251-1260.
Nathubhai, A.; Wood, P. J.; Haikarainen, T.; Hayward, P. C.;
Muñoz-Descalzo, S.; Thompson, A. S.; Lloyd, M. D.; Lehtiö, L.;
Threadgill, M. D. Structure-Activity Relationships of 2-aryl-
quinazolin-4-ones as highly selective and potent inhibitors of
tankyrases. Eur. J. Med. Chem., 2016, 118, 316-327.
Shultz, M. D.; Cheung, A. K.; Kirby, C. A.; Firestone, B.; Fan, J.;
Chen, C. H.-T.; Chen, Z.; Chin, D. N.; DiPietro, L.; Fazal, A.;
Feng, Y.; Fortin, P. D.; Gould, T.; Lagu, B.; Lei, H.; Lenoir, F.;
Venkannagari, H.; Fallarero, A.; Feijs, K. L. H.; Lüscher, B.;
Lehtiö, L. Activity-based assay for human mono-ADP-
ribosyltransferase ARTD7/PARP15 and ARTD10/PARP10 aimed
at screening and profiling inhibitors. Eur. J. Pharm. Sci. 2013, 49,
148-156.
Table of Contents graphic
7
ACS Paragon Plus Environment