Design and discovery of 3-aryl-5-substituted-isoquinolin-1-ones as potent tankyrase inhibitors

Article abstract of DOI:10.1039/c5md00210a

The tankyrase proteins (TNKS, TNKS2), members of the PARP superfamily of enzymes, are attractive anti-cancer drug targets, particularly as inhibition of their catalytic activity has been shown to antagonise oncogenic WNT signalling. To identify chemical inhibitors of tankyrase we carried out an in silico small molecule screen using a set of 'PARP-binding' pharmacophores together with a generated (liganded) tankyrase homology model. This approach identified a structurally diverse set of ～1000 compounds for further study. Subsequent in vitro screening of recombinant tankyrase protein identified a subset of 59 confirmed inhibitors. Early optimisation followed by cell-based studies in WNT-dependent tumour cells, as well as co-crystallisation studies, identified a novel class of 3-aryl-5-substituted isoquinolin-1-ones, such as 21, that exhibit potent inhibition of tankyrase activity as well as growth inhibition of colorectal cancer cells.

Full text of DOI:10.1039/c5md00210a

View Article Online
View Journal
MedChemComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: R. J.R Elliott, A.
Jarvis, M. B. Rajasekaran, M. Menon, L. Bowers, R. Boffey, M. Bayford, S. Firth-Clark, R. Key, R. Aqil, S. B.
Kirton, D. Niculescu-Duvaz, L. Fish, F. Lopes, R. McLeary, I. Trindade, E. Vendrell, F. Munkonge, R. Porter,
T. Perrior, C. Springer, A. Oliver, L. Pearl, A. Ashworth and C. J. Lord, Med. Chem. Commun., 2015, DOI:
10.1039/C5MD00210A.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/medchemcomm

Page 1 of 11
MedChemComm
View Article Online
DOI: 10.1039/C5MD00210A
Lord, Christopher J.; The Institute of Cancer Research, The Breakthrough
Breast Cancer Research Centre and CRUK Gene Function Laboratory

MedChemComm
Page 2 of 11
View Article Online
DOI: 10.1039/C5MD00210A
Design and discovery of 3-aryl-5-substituted-isoquinolin-1-ones as potent tankyrase
inhibitors
Richard J.R. Elliott1‡, Ashley Jarvis2‡, Mohan B. Rajasekaran3‡, Malini Menon1‡, Leandra Bowers1‡, Ray Boffey2, Melanie Bayford2, Stuart
Firth-Clark2, Rebekah Key2, Rehan Aqil2, Stewart B. Kirton§, Dan Niculescu-Duvaz4, Laura Fish4, Filipa Lopes4, Robert McLeary4, Ines Trin-
dade1, Elisenda Vendrell1, Felix Munkonge1, Rod Porter5, Trevor Perrior2, Caroline Springer4, Antony W. Oliver3, Laurence H. Pearl3, Alan
Ashworth1* and Christopher J. Lord1*
1
The Breakthrough Breast Cancer Research Centre and CRUK Gene Function Laboratory, The Institute of Cancer Research, London, UK; 2
Domainex, 162 Cambridge Science Park, Milton Road, Cambridge, UK; 3 CRUK DNA Repair Enzymes Group, Genome Damage and Stability
4
Centre, University of Sussex, UK;
CRUK Centre for Cancer Therapeutics, The Institute of Cancer Research, London, UK;
5RodPorterConsultancy, Ashwell, Herts, UK; §Department of Pharmacy, University of Hertfordshire, Herts, UK. * (C.L. & A.A.) The Institute
of Cancer Research, 237 Fulham Road, London , SW3 6JB, UK. E-mail: Chris.Lord@icr.ac.uk, Alan.Ashworth@icr.ac.uk.

Page 3 of 11
MedChemComm
View Article Online
DOI: 10.1039/C5MD00210A
Abstract:
The tankyrase proteins (TNKS, TNKS2), members of the PARP superfamily of enzymes, are attractive anti-cancer drug targets, particularly
as inhibition of their catalytic activity has been shown to antagonise oncogenic WNT signalling. To identify chemical inhibitors of tanky-
rase we carried out an in silico small molecule screen using a set of ‘PARP-binding’ pharmacophores together with a generated (liganded)
tankyrase homology model. This approach identified a structurally diverse set of ~1000 compounds for further study. Subsequent in vitro
screening of recombinant tankyrase protein identified a subset of 59 confirmed inhibitors. Early optimisation followed by cell-based stud-
ies in WNT-dependent tumour cells, as well as co-crystallisation studies, identified a novel class of 3-aryl-5-substituted isoquinolin-1-
ones, such as 21, that exhibit potent inhibition of tankyrase activity as well as growth inhibition of colorectal cancer cells.

MedChemComm
Page 4 of 11
View Article Online
DOI: 10.1039/C5MD00210A
Introduction
The ADP-ribosyltranferase diphtheria toxin-like (ARTD) or poly-ADP-ribose polymerase (PARP) protein superfamily comprises 17 proteins
that contain a common catalytic domain. Those that are catalytically active use β-NAD+ as an essential co-factor to transfer poly- or
mono-ADP-ribose units onto protein substrates1. This post-translational modification is best characterised for PARP1 (ARTD1) and PARP2
(ARTD2) substrates, which play an important role in the DNA damage response. The PARP1/2 inhibitor olaparib has now been approved
for use in the treatment of ovarian cancer, as it is able to selectively target tumour cells with either a BRCA1 or BRCA2 tumour suppressor
gene defect2. PARP1/2 inhibitors such as olaparib 1 and veliparib 2 exploit the nicotinamidyl pharmacophore present in β-NAD+ 3 (Sup-
plementary Figure 1). Tankyrase 1 and 2 (TNKS/ARTD5 & TNKS2/ARTD6) are PARP proteins which are involved in a range of cellular func-
tions including telomere maintenance, control of the mitotic checkpoint and WNT signalling, as well as the genetic disorder Cherubism3.
The role of TNKS/TNKS2 in WNT signalling has highlighted the potential of tankyrase inhibitors as anti-cancer agents, as this pathway has
been associated with the development of many tumour types, including colorectal cancers, where loss-of-function mutations in the tu-
mour suppressor gene APC4 cause constitutive WNT signalling driving a tumourigenic phenotype. APC acts as a molecular scaffold, or
hub, for the assembly of the ‘destruction complex’, a key component of canonical WNT signalling; this complex, which includes GSK3β
and AXIN1/2, sequesters cytosolic β-catenin (a known transcriptional co-activator). The subsequent phosphorylation of β-catenin, by
GSK3β, promotes its degradation by the proteasome, leading to the inactivation of WNT signalling3.
Conversely, loss of normal destruction complex function, which can occur when APC is mutated, leads to an enhanced level of nuclear,
non-phosphorylated, ‘active’ β-catenin which then drives the transcription of WNT target genes such as MYC5. TNKS and TNKS2 normally
PARylate two components of the destruction complex, AXIN1 and AXIN2, thereby promoting their ubiquitylation and proteosomal deg-
radation; events that minimise the total amount of active β-catenin6. Inhibition of TNKS/TNKS2 minimises AXIN degradation, stabilises
the destruction complex and suppresses WNT signalling6. As constitutive WNT signalling can often be oncogenic, chemical inhibitors of
tankyrase activity have been proposed as potential anti-cancer agents6.
Here we describe the design and discovery of a series of novel 3-aryl-5-substituted isoquinolin-1-one compounds that are potent
TNKS/TNKS2 inhibitors. The optimised compounds presented in this study were characterised by their biochemical potency (estimated
using an in vitro poly-ADP-ribosylation assay), their ability to inhibit cellular WNT signalling (quantified using a transcriptional reporter
assay) and their ability to inhibit growth of WNT-dependent, APC-mutant (APCmut) colorectal tumour cells.
Results and discussion
In silico and biochemical screening for TNKS inhibitors. In order to initiate the discovery of small molecule TNKS/TNKS2 inhibitors, we
used an in silico small molecule screen (Supplementary Figure 2) to identify a set of commercially available compounds that contained
the core (nicotinamide-like) aryl-CONH unit found in PARP superfamily inhibitors such as olaparib 1 (AstraZeneca/KuDOS) and veliparib 2
(AbbVie) (Supplementary Figure 1). One thousand and sixty nine compounds were identified using this approach, which were then as-
sessed using a cell free biochemical assay, where the ability to inhibit PARylation of a histone pseudo-substrate by recombinant TNKS
was quantified7, 8. This analysis produced 59 hits with >75% inhibition of TNKS activity at a concentration of 10 ꢀM (Supplementary Fig-
ure 3). Consequently, several robust chemical series were identified after dose response validation assays (data not shown) of which one,
based on a dihydroisoquinolin-1-one (DHIQ) scaffold, was selected for further optimisation and derivatisation (for an example from the
DHIQ series, Compound 4, see Supplementary Figure 1).
Optimisation of hits from primary screening. At the very start of our studies, there were no confirmed tankyrase inhibitors, or publically
available X-ray crystal structures of ligand-bound protein, so our initial optimisation of the DHIQ series depended on insights gleaned
from in silico molecular docking studies (using a TNKS homology model that we had generated for our initial virtual screen; see Supple-
mentary Information for description) together with structure-activity relationships (SAR) from our other hit series (data not shown).
Shortly after the instigation of our studies, a structurally-related compound, XAV939 5 (Supplementary Figure 1), was shown to antago-
nise WNT signalling by stabilising AXIN1/2 protein levels and was confirmed as a TNKS inhibitor6, therefore we could also benefit from
information contained in this chemical structure. In our design we sought to preserve the classical PARP1 binding motif in the nicotina-
mide-binding region [i.e. the H-bonding network between the lactam carbonyl O with Ser1221 (OH) and Gly1185 (NH) and also the lac-
tam NH with Gly1185 (C=O)]. We also wanted to make use of the rich potential for pi-stacking and non-polar (hydrophobic) interactions
found in the binding site in order to maximise the affinity of our compounds. Analysing how the DHIQ scaffold docked into the TNKS ho-
mology model allowed us to consider ‘morphing’ from DHIQ into a flattened isoquinolin-1-one (IQ) scaffold thereby optimising the pi-
stacking interaction with Tyr1224. We further envisioned aryl substitution of the lactam-containing ring (XAV939 5 contains an aryl-
substituted lactam), with scope for growth via the para- position. We were aware that the isoquinolin-1-one scaffold had featured in
PARP superfamily inhibitors, with examples that incorporated 5-substitution (Compound 4 possesses a 5-substituted –OCH2CONH-aryl
motif; Supplementary Figure 1), but our planned 3-aryl substitution, and subsequent para-growth, afforded a unique, differentiating
opportunity. Finally, we sought to replace the ‘inefficient’ –OCH2CONH-aryl motif incorporated in 4 (Supplementary Figure 1) with a
smaller group envisaged to be more compatible with the target protein. Our model for the proposed binding of an optimised IQ exam-
ple, as represented by compound 11, is shown in Figure 1A.
Structure-based drug design. Again, using docking to guide our design efforts, we prepared and tested a set of 3-aryl-isoquinolin-1-one
compounds, which upon further optimisation, led to a preferred series of 3-aryl-5-substituted-isoquinolin-1-ones (Table 1) that showed

Page 5 of 11
MedChemComm
View Article Online
DOI: 10.1039/C5MD00210A
improved TNKS potency, especially when compared to 4. Recently, a range of structurally diverse TNKS inhibitors have been report-
ed9,10,11,12. In 2012, an X-ray crystal structure of XAV939 5 bound to the nicotinamide-binding pocket of the catalytic domain of TNKS13
reinforced the key features of the IQ binding model that we had envisaged (Figure 1B). Thus, with a now established and robust struc-
ture-based paradigm we turned our attention to progression of our hit compounds, designing a test cascade — beginning with the gen-
eration of dose response curves using recombinant TNKS in PARylation assays, then moving onto a cell-based WNT (LEF/TCF) luciferase
reporter system that assessed ability to inhibit β-catenin driven mRNA transcription — to select promising compounds that were then
assessed for their selectivity against PARP1.
Biological assessment of optimised compounds. The bioactivity SAR (Table 1) highlighted an important role for the R1 group, with exam-
ples containing R1 = OMe (compounds 10, 11, 12) showing enhanced inhibition of the WNT-luciferase (WNT-Luc) signal, compared with
R1 = H, F and Cl (compounds 13, 14, 15 and 16). Compound 11, for example, not only showed good TNKS inhibitory activity but also
demonstrated excellent selectivity over full length PARP1 (IC50 7900 nM, >200-fold) and in DLD1 human colorectal tumour cells, inhibited
WNT-reporter gene transcription with an IC50 = 59 nM. Unfortunately, other compound properties (where determined), such as aqueous
solubility and microsomal stability, required improvement, although we found that the solubility could be greatly enhanced with the
incorporation of an amine at R2 (e.g. > 100 ꢀM in phosphate-buffered saline at pH 7.4 for compounds 12 and 16 with R2 = -CH2NMe2).
As one of our future aspirations was to be able to explore the pharmacological potential for tankyrase inhibitors using murine models of
human cancer, we recognised that we would need to identify compounds that possessed an acceptable balance between potency and
pharmacokinetic (PK) properties. Encouraged by our initial results, we prepared and tested a focused group of IQ compounds (Table 2)
that combined R1 = Me, included as a surrogate for the potent, but metabolically-labile, OMe group (O-demethylation was confirmed by
metabolite identification using LC-MS), with R2 = CH2’N’ (e.g. CH2NMe2). These compounds retained or enhanced the desired mechanistic
(WNT pathway) cellular potency and were able to inhibit growth of an APC-mutant colorectal cell line, DLD1 (Table 2). In support of this
observation, a recent report has indicated that a methyl group occupying a comparable (8-) position on a series of 2-arylquinazoline-4-
one TNKS/TNKS2 inhibitor also enhances potency14. Furthermore, additional improvements were apparent, e.g. compound 21 had good
aqueous solubility (> 100 ꢀM in phosphate-buffered saline at pH 7.4) and demonstrated good metabolic stability upon exposure to
mouse and human liver microsomes (MLM/HLM Clint = 10/5 ꢀL/min/mg protein). Subsequent to this, using X-ray crystallographic studies,
we sought to reinforce our understanding of the structural underpinnings for compound potency, and were able to determine the struc-
ture of compound 12 bound to TNKS at 3.4 Ångström resolution (see Supplementary Information for additional details). The binding
mode of 12, determined by crystallography, was comparable to that observed for XAV939 5 and also to the optimum molecular docking
pose determined with our TNKS homology model and compound 11 (Figure 1A, B, C). Furthermore, we then went on to determine crys-
tal structures of 21 and 23 in complex with TNKS, at 2.4 and 2.5 Ångström respectively (Figure 1D, E).
Each of the three compounds made the same core set of hydrogen-bonds, i.e. from the NH and carbonyl groups of the isoquinolin-1-one
scaffold, to the backbone carbonyl and NH of Gly1185, respectively. An additional interaction was also made to the hydroxyl group of
the Ser1221 side chain. Furthermore, Tyr1203 of the ‘D-loop’ motif (amino acids Phe1197-Gly1211 – subsequently targeted as one of the
primary areas to influence selectivity vs. PARP1,215) as well as Tyr1213 and Tyr1224, were involved in van der Waals interaction with the
bound compound; in particular Tyr1224 was pi-stacked up against the face of the isoquinolin-1-one core, as predicted from our original
molecular modeling study (Figure 1A). When the 12, 21 and 23 co-crystal structures were overlaid, small movements in the position of
the Phe1188 side chain were also evident (Figure 1F); this was concomitant with a van der Waals interactions of this side chain with the
additional pendant rings of 21 and 23, when compared to 12, and may therefore provide a favorable interaction which contributes to the
increased potency of these compounds. It is worth noting here that our TNKS homology model was built with the structural data availa-
ble at the time, and so our docking poses did not fully account for the conformational flexibility of the D-loop region, and therefore did
not predict the interaction of bound ligands with the side chain of, e.g. Tyr1203. This, however, did not seem to affect the validity of the
docking poses generated.
Compounds were also assessed for their ability to stabilise both tankyrase and AXIN1/2 protein levels (by inhibiting auto-PARylation and
substrate PARylation, respectively) and destabilise β-catenin (for example, Figure 2A, compound 21). Upon testing 21 in a larger panel of
tumour cell lines (Figure 2B), preferential potency against APC-mutant lines compared to non-APC-mutant lines was observed, suggesting
that 21 would appear to be representative of a potent class of TNKS inhibitor, with favorable tumour cell inhibitory properties.
Whilst these cell inhibitory properties of 21 might only be partially due to TNKS inhibition, and possibly due to PARP1 or other PARP su-
perfamily inhibitory effects (for example, the selectivity of 21 for TNKS over PARP1 in a cell free biochemical assay was only 35-fold), we
noted that Compound 21 only impaired cellular DNA-damage induced (PARP1-mediated) PARylation at concentrations > 100 nM (Sup-
plementary Figure 4). Furthermore, the PARP inhibitor olaparib 1 showed particularly poor activity in our WNT-luciferase assay (IC50
3534 nM) and did not alter protein levels of AXIN1/2 or β-catenin and only affected TNKS stability at high concentrations (Supplementary
Figure 5A, B). Cell lines expressing mutant APC were unaffected by olaparib 1 at concentrations that normally inhibit BRCA1 or BRCA2
mutant tumour cells (DLD1 SF50 2000 nM, HT55 SF50 6265 nM). So, whilst selectivity of a TNKS inhibitor against PARP1 may be desirable
per se, any PARP1 inhibitory (off-target) effect is unlikely to have an impact on cell viability in APC-mutant colorectal cancer. However,
since compound 21 was not assayed against the remaining PARP family members we cannot preclude the effect that complete or partial
inhibition of these related enzymes may have on the viability of the cell line models tested.

MedChemComm
Page 6 of 11
View Article Online
DOI: 10.1039/C5MD00210A
Experimental
The synthesis of 3-aryl-5-methyl isoquinolin-1-one compounds 10-23 (see Supplementary Information for full experimental details) was
achieved by the one-pot, lithiated toluamide-aryl nitrile cycloaddition method16, 17 using toluamides 7a–7e and 4-(R2)-aryl nitriles (see
Scheme 1 for representative examples 11, 12, 17, 19, 21, 23). The toluamides were either commercially available 7e or prepared using a
coupling reaction between substituted benzoic acids 6a-6d and diethylamine. The 4-(R2)-aryl nitriles 9a-9b were prepared in moderate
yield using a halide displacement reaction between 4-bromomethyl-benzonitrile 8 and the corresponding, commercially available sec-
ondary amines. The remaining 4-(R2)-aryl nitriles (where R2 = F, Me, CH2NMe2, CF3, CH2-N-morpholine, CH2-N-(4-Me)-piperazine, CH2-N-
(4-Boc)-piperazine) were all commercially available.
Conclusions
In summary, we have described the discovery, optimisation and biological properties of a novel series of 3-aryl-5-substituted-isoquinolin-
1-ones as potent inhibitors of the PARP superfamily enzyme, TNKS. In particular, a class of 3-aryl-5-methyl-isoquinolin-1-one compounds
demonstrated enhanced potency in biochemical and cell-based assays as well as improved solubility and metabolic stability.
Notes and references:
For complete experimental details for the TNKS homology model generation and virtual screen, chemical synthesis, biological assays and
X-ray crystallography see Supplementary Information. The PDB accession codes for the X-ray co-crystal structure of TNKS + 12, TNKS +
21 and TNKS + 23 are 4UW1, 4U6A and 4UUH, respectively.
Acknowledgements
This work was funded by the Wellcome Trust as part of its Seeding Drug Discovery Initiative (SDDI, WT101816). The funders took no part
in the drafting of this manuscript. The licensing or sale of intellectual property described in this paper may result in income that will be
shared by the ICR and those who contributed directly to the intellectual property under the Institutes 'Rewards to Discovers Scheme'.
Part of this work has been described in a PCT application published as WO2013/132253 A1.
1
2
3
4
M. Rouleau, A. Patel, M. J. Hendzel, S. H. Kaufmann and G. G. Poirier, Nat Rev Cancer, 2010, 10, 293-301.
C. J. Lord and A. Ashworth, Nature, 2012, 481, 287-294.
J. L. Riffell, C. J. Lord and A. Ashworth, Nature reviews. Drug discovery, 2012, 11, 923-936.
S. M. Powell, N. Zilz, Y. Beazer-Barclay, T. M. Bryan, S. R. Hamilton, S. N. Thibodeau, B. Vogelstein and K. W. Kinzler, Nature,
1992, 359, 235-237.
5
6
T. C. He, A. B. Sparks, C. Rago, H. Hermeking, L. Zawel, L. T. da Costa, P. J. Morin, B. Vogelstein and K. W. Kinzler, Science, 1998,
281, 1509-1512.
S. M. Huang, Y. M. Mishina, S. Liu, A. Cheung, F. Stegmeier, G. A. Michaud, O. Charlat, E. Wiellette, Y. Zhang, S. Wiessner, M.
Hild, X. Shi, C. J. Wilson, C. Mickanin, V. Myer, A. Fazal, R. Tomlinson, F. Serluca, W. Shao, H. Cheng, M. Shultz, C. Rau, M.
Schirle, J. Schlegl, S. Ghidelli, S. Fawell, C. Lu, D. Curtis, M. W. Kirschner, C. Lengauer, P. M. Finan, J. A. Tallarico, T.
Bouwmeester, J. A. Porter, A. Bauer and F. Cong, Nature, 2009, 461, 614-620.
7
8
C. C. Clark, J. N. Weitzel and T. R. O'Connor, Mol Cancer Ther, 2012, 11, 1948-1958.
M. Fraser, H. Zhao, K. R. Luoto, C. Lundin, C. Coackley, N. Chan, A. M. Joshua, T. A. Bismar, A. Evans, T. Helleday and R. G.
Bristow, Clin Cancer Res, 2012, 18, 1015-1027.
9
A. Voronkov, D. D. Holsworth, J. Waaler, S. R. Wilson, B. Ekblad, H. Perdreau-Dahl, H. Dinh, G. Drewes, C. Hopf, J. P. Morth and
S. Krauss, Journal of medicinal chemistry, 2013, 56, 3012-3023.
10
H. Bregman, N. Chakka, A. Guzman-Perez, H. Gunaydin, Y. Gu, X. Huang, V. Berry, J. Liu, Y. Teffera, L. Huang, B. Egge, E. L.
Mullady, S. Schneider, P. S. Andrews, A. Mishra, J. Newcomb, R. Serafino, C. A. Strathdee, S. M. Turci, C. Wilson and E. F.
DiMauro, Journal of medicinal chemistry, 2013, 56, 4320-4342.
11
M. D. Shultz, A. K. Cheung, C. A. Kirby, B. Firestone, J. Fan, C. H. Chen, Z. Chen, D. N. Chin, L. Dipietro, A. Fazal, Y. Feng, P. D.
Fortin, T. Gould, B. Lagu, H. Lei, F. Lenoir, D. Majumdar, E. Ochala, M. G. Palermo, L. Pham, M. Pu, T. Smith, T. Stams, R. C.
Tomlinson, B. B. Toure, M. Visser, R. M. Wang, N. J. Waters and W. Shao, Journal of medicinal chemistry, 2013, 56, 6495-6511.
P. Liscio, A. Carotti, S. Asciutti, T. Karlberg, D. Bellocchi, L. Llacuna, A. Macchiarulo, S. A. Aaronson, H. Schuler, R. Pellicciari and
E. Camaioni, Journal of medicinal chemistry, 2014, 57, 2807-2812.
12
13
C. A. Kirby, A. Cheung, A. Fazal, M. D. Shultz and T. Stams, Acta crystallographica. Section F, Structural biology and
crystallization communications, 2012, 68, 115-118.
14
15
A. Nathubhai, P. J. Wood, M. D. Lloyd, A. S. Thompson and M. D. Threadgill, Acs Med Chem Lett, 2013, 4, 1173-1177.
E. Wahlberg, T. Karlberg, E. Kouznetsova, N. Markova, A. Macchiarulo, A-G. Thorsell, E. Pol, A. Frostell, T. Ekblad, D. Öncü, B.
Kull, G. M. Robertson, R. Pellicciari, H. Schüler and J. Weigelt, Nature Biotechnology, 2012, 30, 283-288.
W. J. Cho, S. J. Yoo, M. J. Park, B. H. Chung and C. O. Lee, Archives of pharmacal research, 1997, 20, 264-268.
G. S. Poindexter, J Org Chem, 1982, 47, 3787-3788.
16
17

Page 7 of 11
MedChemComm
View Article Online
DOI: 10.1039/C5MD00210A
Figure 1

MedChemComm
Page 8 of 11
View Article Online
DOI: 10.1039/C5MD00210A
Figure 2

Page 9 of 11
MedChemComm
View Article Online
DOI: 10.1039/C5MD00210A
Scheme 1.
O
O
OH
NEt2
a
R1
R1
6a: R1 = OMe
6b: R1 = Cl
6c: R1 = F
7a: R1 = OMe
7b: R1 = Cl
7c: R1 = F
6d: R1 = Me
7d: R1 = Me
N
N
b
R2
Br
9a: R2 = CH2-N-(4-NMe2)-piperidine
9b: R2 = CH2-N-(3-NMe2)-azetidine
8
O
O
NEt2
NH
c
R1
R1
R2
Compounds 10 - 23, e.g.:
7a: R1 = OMe
7b: R1 = Cl
7c: R1 = F
7d: R1 = Me
7e: R1 = H
11: R1 = OMe, R2 = Me
12: R1 = OMe, R2 = CH2NMe2
17: R1 = Me, R2 = CH2NMe2
19: R1 = Me, R2 = CH2-N-(4-Boc)-piperazine
23: R1 = Me, R2 = CH2-N-piperazine
21: R1 = Me, R2 = CH2-N-(4-NMe2)-piperidine

MedChemComm
Page 10 of 11
View Article Online
DOI: 10.1039/C5MD00210A
Figure Legends
Figure 1. (A) Representative docking pose for IQ compound 11 bound to the homology model of TNKS.
(B) Predicted binding mode of 11 compared to that observed for XAV939 5. (C) Confirmation of the
binding mode as determined by an X-ray co-crystal structure of 12 in complex with TNKS. (D) TNKS in
complex with 21. (E) TNKS in complex with 23. (F) Overlay of each crystal structure, highlighting the
movement of Phe1188 in each case (van der Waals radii represented by dotted surface). Potential hy-
drogen bonds to key amino acids (as labeled) are shown as orange dashed lines.
Figure 2. (A) Western blot illustrating effect on TNKS, AXIN2, β-catenin and Tubulin protein levels in
SW480 colorectal tumour cells exposed to 21. (B) Cell survival plots illustrating data from multiple colo-
rectal tumour cell lines exposed to 21. Cell lines with mutant APC alleles are indicated in black, and
those with wild-type APC in blue. Error bars represent SEM from ≥3 replica experiments.
Scheme 1. Synthesis of Compounds 10-23. Reagents and conditions: (a) Et2NH, TBTU, DIPEA,
DCM/DMF (1:1), rt or (COCl)2, DCM, cat. DMF then Et2NH, DCM, 0 oC to rt ; (b) 4-
o
(Dimethylamino)piperidine, Et3N, THF, 50 C or 3-(dimethylamino)azetidine dihydrochloride, K2CO3,
o
o
o
MeCN, rt (c) i. n-BuLi, THF, -78 C, ii. 4-(R2)-ArCN, THF, -78 C or i. n-BuLi, THF, -78 C, ii. 4-(R2)-ArCN,
THF, -78 oC, iii. TFA/DCM (1:2), rt.

Page 11 of 11
MedChemComm
View Article Online
DOI: 10.1039/C5MD00210A
TABLES
Table 1. SAR of 3-aryl-5-substituted isoquinolin-1-ones
TNKS IC50
(nM)
WNT-Luc IC50
(nM)
Compound
R1
R2
TNKS 95% CI (nM)
WNT-Luc 95% CI (nM)
10
11
12
13
14
15
16
Methoxy
Methoxy
Methoxy
Chloro
Chloro
Fluoro
H
Fluoro
Methyl
104
34
92 - 117
23 - 50
120
59
77 - 186
50 - 70
CH2NMe2
Trifluoromethyl
CH2NMe2
CH2NMe2
CH2NMe2
19
17 - 23
136
108 - 170
313 - 744
235 - 566
621 - 4790
Not applicable
128
151
16
101 - 162
136 - 167
13 - 19
483
364
1730
>10000
18
16 - 21

MedChemComm
Page 12 of 11
View Article Online
DOI: 10.1039/C5MD00210A
Table 2. SAR of 3-aryl-5-methyl isoquinolin-1-ones
TNKS
IC50
(nM)
TNKS
95% CI
(nM)
PARP1
IC50
(nM)
PARP1
95% CI
(nM)
WNT-Luc
95% CI
(nM)
APCmut
DLD1 95%
SF50 DLD1
CI (nM)
WNT-Luc
IC50 (nM)
Compound
17
R2
(nM)*
12
9-16
267
224-318
72
37-140
1000
554-2040
18
20
12
17
10-15
15-18
ND
ND
25
43
23-28
27-68
ND
ND
652
470-905
886
645-1220
21
22
13
21
9-20
465
ND
376-576
ND
61
35
51-75
27-46
80
59 - 108
ND
14-30
ND
23
17
12-24
ND
ND
117
73-170
2521
1150- 5526
* ND = Not determined