Tankyrase Inhibition Enhances the Anti-proliferative Effect of PI3K and
EGFR Inhibition, Mutually Affecting E-catenin and AKT Signaling in
Colorectal Cancer

Authors and affiliations
Nina T. Solberg, Jo Waaler, Kaja Lund, Line Mygland, Petter A. Olsen, Stefan Krauss
Affiliation:
Department of Microbiology, Section for Cell Signaling, Oslo University Hospital￾Rikshospitalet, Oslo 0373, Norway
and
Hybrid Technology Hub – Centre of Excellence, Institute of Basic Medical Sciences,
Faculty of Medicine, University of Oslo, PO Box 1112 Blindern, Oslo 0317, Norway
Running title
Crosstalk Between Tankyrase, PI3K and EGFR inhibitors in CRC
Keywords
Tankyrase inhibition, PI3K inhibition, EGFR inhibition, G007-LK, CRC
Additional information
Financial support: The work was supported by grants given to Stefan Krauss by the
Norwegian Cancer Society contract 2327614, 4561000 and 5803958, the Research
Council of Norway contract 226290/030 and the Norwegian Health Region South
East contract 2015012.
Corresponding author:
Nina Therese Solberg, Unit for Cell Signaling, Department of Microbiology, Oslo
University Hospital-Rikshospitalet, Domus Medica II- Plan2 – rom L-208,
Gaustadalleen 34, 0372 Oslo, Norway
Email: [email protected]
Phone: +47 23 07 90 14/ +47 93 88 34 84
Conflict of interest: G007-LK development is described in Voronkov et al., 2013
(reference #18) and the compound is patented (Stefan Krauss and Jo Waaler). No
potential conflicts of interest were disclosed by the other authors.
Word count: 4980
Number of figures and tables: 7 (21 supplementary)
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
2
ABSTRACT
Over-activation of the WNT/β-catenin signaling axis is a common denominator in
colorectal cancer (CRC). Currently there is no available WNT inhibitor in clinical
practice. Although tankyrase inhibitors have been proposed as promising candidates
there are many CRC models that do not respond positively to tankyrase inhibition in
vitro and in vivo. Therefore, a combinatorial therapeutic approach combining a
tankyrase inhibitor (G007-LK) with PI3K (BKM120) and EGFR (Erlotinib) inhibitors
in CRC was investigated. The data demonstrate that tankyrase inhibition enhances the
effect of PI3K and EGFR inhibition in the tankyrase inhibitor sensitive
COLO320DM, and in the non-sensitive HCT-15 cell line. In both cell lines, combined
tankyrase/PI3K/EGFR inhibition is more effective at reducing growth than a dual
tankyrase/MEK inhibition. Tankyrase/PI3K/EGFR inhibition affected in a context￾dependent manner components of the WNT/E-catenin, AKT/mTOR, EGFR and RAS
signaling pathways. Tankyrase/PI3K/EGFR inhibition also efficiently reduced growth
of both COLO320DM and HCT-15 tumor xenografts in vivo. At the highest doses
tumor xenograft growth was halted without affecting the body weight of the tested
animals.
Implications: Combining tankyrase inhibitors with PI3K and EGFR inhibition may
expand the therapeutic arsenal against colorectal cancers.
INTRODUCTION
Colorectal cancers (CRC) are some of the most common cancers worldwide,
and are frequently initiated by mutations in the adenomatous polyposis coli (APC)
tumor suppressor gene or the gene encoding β-catenin (CTNNB1) (1). Subsequent
mutations in KRAS and TP53 and deregulation of signaling pathways like
phosphatidylinositol-3-kinase-(PI3K)/protein kinaseB (AKT) and transforming￾growth factor-beta (TGFβ), are further hallmarks for CRC development and
progression (2,3).
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
3
Targeted therapy against a number of molecular targets is being explored for
treatment of CRC. Epidermal growth factor receptor (EGFR) is an upstream regulator
of two central pathways; the mitogen-activated protein kinase (MAPK) pathway and
the PI3K/AKT pathway. Both pathways regulate cellular proliferation, migration,
differentiation and apoptosis (4,5). Several inhibitors targeting EGFR have been
developed, including Erlotinib (6) and Gefitinib (7), and a number of PI3K inhibitors
are in clinical studies (8,9). Tankyrase is a central cytoplasmic biotarget in the
WNT/β-catenin signaling pathway where it controls the turnover of AXIN1/2, and
thereby prevents degradation of β-catenin (10-12). Experimental tankyrase inhibitors
such as XAV939 and G007-LK have gained increasing attention as inhibitors for
WNT induced CRC (10,12). However, despite long term tolerability of tankyrase
treatment in mice (13), the advance of tankyrase inhibitors to clinical trials is at
current hampered by cytotoxicity issues (12,14).
Multiple interactions have been mapped between the WNT, PI3K/AKT and
EGFR signaling pathways. In lung cancer cells tankyrase activity was shown to
protect cells from EGFR inhibition, an effect that was counteracted by combined
tankyrase/EGFR inhibition (15). In CRC, high levels of nuclear β-catenin confer
resistance to apoptosis-inducing signals mediated by PI3K/AKT inhibition, an effect
that can be reversed by exposure to tankyrase inhibition (16). Furthermore, depending
on the mutational background, EGFR, PI3K in addition to MEK inhibitors have
shown to enhance the effect of tankyrase inhibitors in CRC cell lines (17).
In this study we use the highly specific tankyrase1/2 inhibitor G007-LK (18),
the pan-class-I PI3K inhibitor BKM120 (8), and the EGFR inhibitor Erlotinib (6) to
investigate the effect of single and combined inhibition in two representative CRC
cell lines. Using tankyrase inhibitor sensitive COLO320DM, and in-sensitive HCT-
15, we show that tankyrase inhibition can be potentiated by EGFR and PI3K
inhibition both in vitro and in vivo. At the tested doses, combined
tankyrase/PI3K/EGFR inhibition was more efficacious than combined
tankyrase/MEK inhibition in vitro.

Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
4
MATERIALS AND METHODS
Cell culture
COLO320DM, COLO205, HCT-8 and HCT-15 cells were cultured in RPMI-1640
(Sigma-Aldrich), WiDr and RKO in Eagle’s Minimum Essential Medium (LGC￾standards), containing 10% FBS, and 1% penicillin/streptomycin at 37o
C and 5%
CO2. SW403 and SW480 cells were cultured in Leibovitz’s L-15 with L-glutamine
(Sigma-Aldrich), 10% FBS, and 1% penicillin/streptomycin at 37o
C and 0% CO2. Cell
lines were obtained from ATCC, and routinely tested for Mycoplasma using the
MycoAlert Mycoplasma detection kit (Lonza). Cell line authentication was performed
on COLO320DM, HCT-15, SW403 and SW480 cell lines.
Inhibitors
Inhibitors were dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich), which was
also used as vehicle control. The following inhibitors were used: G007-LK
(ChemRoyal), BKM120 (NVP-BKM120; Chemietek), API-2 (TOCRIS), and GDC-
0941 (Selleckchem), Erlotinib (Chemietek), Gefitinib (Selleckchem) and GDC-0973
(MedChem Express).
IncuCyte proliferation assay
1000 COLO320DM or HCT-15 cells, 7000 SW403 cells, or 1200 SW480 cells were
seeded in 96-well plates (minimum 3 replicates). The next day, culture media was
replaced with drug containing media. Cell confluence was quantified with the
IncuCyte live-cell analysis system (Essen BioScience).
Colony assay
400 COLO320DM, HCT-15 or SW480 cells, or 50,000 SW403 cells, were seeded in
6-cm plates (6 technical replicates) in growth media containing 1% FBS. Culture
media was replaced with drug containing media the next day, and further twice per
week for 2-3 weeks. Colonies were stained and fixed with 0.2% methylene blue in
methanol and counted.

Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
5
Cell cycle analysis
After 72 hours of treatment, cells were fixed in EtOH for 2 hours at -20o
C, and
propidium iodide (PI) for 30 minutes at RT. For flow cytometric analysis (Attune) at
least 10,000 cells were analysed.
Apoptosis analysis
After 72 hours of treatment, trypsinized cells were incubated in AnnexinV binding
buffer containing 1:500 AnnexinV-FITC Apoptosis Detection Reagent (abcam) and
PI for 5 minutes prior to flow cytometric analysis (Attune).
Immunoblotting
After 72 hours of treatment, total protein extracts were obtained with RIPA buffer
(Millipore), and nuclear/cytoplasmic protein fractions were obtained with NE-PER
Nuclear and Cytoplasmic Extraction Reagents (ThermoFisher). Equal amounts of
protein were denatured, separated on SDS-PAGE gels (Bio-Rad Laboratories) and
transferred to polyvinylidene-difluoride-membranes (Millipore). After blocking in 5%
skim milk (AppliChem)/TBS-T membranes were probed with primary antibody over
night at 4o
C. Following secondary antibody incubation, proteins were visualized with
chemiluminescent substrate (ECL prime Western Blotting Detection Reagent, Sigma￾Aldrich).
esiRNA mediated knock down
800,000 COLO320DM and 500,000 HCT-15 was transfected with 50nM EGFP
(EHUEGFP, Sigma) or TP53 (EHU123221, Sigma) esiRNA using Lipofectamine
RNAiMax Transfection Reagent (ThermoFischer, 13778075) in culture medium
without antibiotics.
RT-qPCR
After 72 hours of treatment, total RNA was obtained using the GenElute miniprep kit
(Sigma Aldrich). cDNA was synthesized using SuperScript® VILO cDNA synthesis
Kit (Life Technologies). RT-qPCR was carried out using a TaqMan Gene expression
Mastermix (Life Technologies) on the StepOnePlus™ PCR Systems (Life
Technologies) and normalized to the GAPDH internal control.
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
6
Immunofluorescence
100,000 COLO320DM and 50,000 HCT-15 cells were plated onto Poly-L-lysine
Solution coated glass covers. The following day, cells were treated for 72 hours, and
subsequently fixed in 4% PFA, permeabilized with 0.1% Triton-X100, incubated with
primary antibodies in 4% BSA for 1 hour, secondary antibody in 4% BSA for 1 hour,
and counterstained with DAPI. Cells were imaged using a Zeiss Elyra PS1
Microscope.
Xenografts
Female CB17SCID mice were implanted subcutaneously with 4,000,000 cells. After
tumor development, animals were randomized, 10 mice per group, and desired
treatment was administered orally as described in details in Supplementary Materials
and Methods. Animal experiments were approved by local ethics authorities at
Norwegian Food Safety Authority (Norway) or Ethics Committee for Animal
Experimentation (Germany) and carried out following accepted ethical standards.
Statistical analysis
SigmaPlot®11 (Systat Software Inc.) was used to perform statistical analyses. P<0.05
was regarded as a statistically significant difference when using Students t-test and
Mann-Whitney rank sum tests.
Gene probes, antibodies and further details are listed in Supplementary Materials and
Methods.
RESULTS
G007-LK enhances the effect of PI3K/AKT and EGFR inhibition on growth
reduction in CRC cells
To investigate the benefit of combining tankyrase inhibition with PI3K and
EGFR signaling inhibition in human CRC, four APC mutant CRC cell lines were
initially selected; COLO320DM and SW403 (tankyrase inhibitor sensitive), HCT-15
and SW480 (tankyrase inhibitor insensitive). Somatic mutations are listed in
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
7
Supplementary Fig. S1. Inhibitor doses providing moderate growth inhibition in
COLO320DM were selected (1 μM G007-LK; 0.5 μM BKM120; 5 PM Erlotinib) and
used on all cell lines (dose response curves in Supplementary Fig. S2. Results
regarding SW403 and SW480 cell lines are reported in Supplementary Fig. S3-S9).
At the selected doses, G007-LK significantly reduced growth in
COLO320DM (Fig. 1A and Supplementary Fig. S10), while HCT-15 cells were
insensitive to single G007-LK treatment (Fig. 1B and Supplementary Fig. S10). Both
BKM120 and Erlotinib significantly reduced growth in both COLO320DM and HCT-
15 cell lines. Combining G007-LK with either BKM120 or Erlotinib further reduced
proliferation in COLO320DM. In HCT-15, G007-LK insignificantly enhanced the
effect of BKM120 and Erlotinib. BKM120 enhanced the effect of Erlotinib treatment
in COLO320DM, while in HCT-15 the effect was highly augmented. Adding G007-
LK to BKM120/Erlotinib further reduced proliferation both in COLO320DM and
HCT-15 cells, leading to a total of 83% and 92% growth reduction compared to
control, respectively. Biotarget specific inhibition was confirmed with alternative
PI3K/AKT (GDC-0941 and API-2, Supplementary Fig. S7) and EGFR (Gefitinib,
Supplementary Fig. S8) inhibitors. The combined treatment effect of G007-
LK/BKM120/Erlotinib was further confirmed in COLO205, HCT-8, WiDr and RKO
CRCs (Supplementary Fig. S11)
MEK inhibition has been identified as a sensitizing factor for tankyrase
inhibition, in particular in KRAS mutant CRC (17) and was therefore compared to the
G007-LK/BKM120/Erlotinib treatment regime. MEK inhibition (1 μM GDC-0973)
slightly (not significant) enhanced growth in KRAS wild type COLO320DM cells, and
moderately (significant) reduced growth in KRAS mutant HCT-15 cells (Fig. 1C-D).
Further, GDC-0973 significantly sensitized HCT-15 cells to G007-LK treatment, and
G007-LK significantly sensitized COLO320DM cells to GDC-0973 treatment.
However, combined G007-LK/BKM120/Erlotinib treatment was more effective on
growth inhibition than combined G007-LK/GDC-0973 treatment at the selected doses
in both cell lines, regardless of the KRAS mutation status (molecular responses in
Supplementary Fig. S12).
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
8
Combination of inhibitors reduces the colony forming ability of CRC and
induces molecular characteristics of differentiation
The colony formation assay is an in vitro cell survival assay based on the
ability of a single cell to grow into a colony. We tested the colony forming ability of
the CRC in the presence of inhibitors in media containing low serum (1% FBS).
COLO320DM cells showed reduced ability to form colonies in response to inhibitor
treatments (not significant with G007-LK), while both G007-LK and BKM120
significantly enhanced Erlotinib mediated colony reduction (Fig. 2A, Supplementary
Fig. S9). In accordance with the proliferation assay, combined G007-
LK/BKM120/Erlotinib treatment reduced colony numbers to 2% compared to control.
In contrast, HCT-15 cells showed increased colony formation with single inhibitor
treatments (significantly with Erlotinib) (Fig. 2B, Supplementary Fig. S9). Combining
G007-LK with BKM120 significantly enhanced the numbers of colonies, while
combining G007-LK with Erlotinib attenuated Erlotinib mediated induction in colony
forming ability. Only combined BKM120/Erlotinib treatment reduced colony
formation compared to the control, and a significant reduction to 35% of the control
was seen with G007-LK/BKM120/Erlotinib treatment.
In both cell lines combined G007-LK/BKM120/Erlotinib treatment most
efficiently reduced both proliferation and colony formation.
Since the colony formation may be influenced by the expression of stem cell
and differentiation markers, we analyzed the expression of leucine-rich repeat￾containing heterotrimeric guanine nucleotide-binding protein-coupled receptor 5
(LGR5), a marker for colon crypt stem cells (19), and the differentiation marker
cytokeratin 20 (KRT20). In COLO320DM, G007-LK induced molecular
characteristics of differentiation by reducing LGR5 and increasing KRT20 expression
(Fig. 2C). BKM120 did not affect expression of either marker, while Erlotinib
enhanced LGR5 and reduced KRT20 expression, supposedly moving the CRC cells
towards a more stem like character. G007-LK counteracted the effect of Erlotinib,
increasing the expression of KRT20 and moderating LGR5 expression.
In HCT-15 cells, G007-LK strongly induced KRT20 and reduced LGR5
expression, while both BKM120 and Erlotinib induced expression of both markers
(Fig. 2D). G007-LK attenuated the BKM120 and Erlotinib induced LGR5 induction,
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
9
and massively increased KRT20 levels (243.5 fold with G007-LK/BKM120/Erlotinib
treatment).
In conclusion, G007-LK most effectively induced molecular characteristics of
differentiation in both cell lines, while Erlotinib promoted stem cell characteristics in
COLO320DM cells. Combined G007-LK/BKM120/Erlotinib treatment was less
efficient on inducing differentiation than single G007-LK treatment. In HCT-15 cells
combined G007-LK/BKM120/Erlotinib attenuated stem cell characteristics induced
by BKM120 and Erlotinib and rather enhanced differentiation.
Changes in proliferation and colony formation are accompanied by changes in
cell cycle genes and apoptosis
To investigate mechanisms responsible for altered proliferation and colony
formation in COLO320DM and HCT-15 cells, flow cytometric analysis of apoptosis
and cell cycle distribution was performed. In addition, we investigated the expression
of the CMYC and cyclinD1 (CCND1) genes which are regulated by a multitude of
factors including the WNT/E-catenin (20) and PI3K/AKT signaling pathways (21,22),
and are required for the G1/S transition (23,24). Induced cMYC levels may also
inhibit expression of cyclin dependent kinase 1A (CDKN1A) and 1B (CDKN1B) genes
(22), which are negative regulators of cell cycle progression from G1 to S phase
through inhibition of CCND1 expression (25).
In COLO320DM, neither G007-LK nor BKM120 alone significantly changed
the number of AnnexinV positive (apoptotic) cells, but when combined apoptosis was
significantly induced compared to control (Fig.3 A). Apoptosis could not be related to
nuclear levels of β-catenin and FOXO3a (16) (Supplementary Fig. S13). Single
Erlotinib treatment significantly induced apoptosis, which was further increased in
combination with BKM120, but not with G007-LK. Combined BKM120/Erlotinib
treatment most potently induced apoptosis (> 3.5 fold), an effect that was attenuated
by G007-LK.
In COLO320DM G007-LK treatment significantly increased the number of
G1 phase cells (Fig. 3B and Supplementary Fig. S14), indicative of cell cycle arrest
and consistent with moderate growth inhibition. This was supported by reduced
expression of CCND1 and CMYC, and increased expression of CDKN1A and
CDKN1B (Fig. 3C and Supplementary Fig. S14). The tendency to more G1 phase
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
10
cells with BKM120 and Erlotinib became statistically relevant when each were
combined with G007-LK. G1 phase increase was accompanied by fewer S phase
cells, reduced expression of CMYC and CCND1, and increased expression of
CDKN1A and CDKN1B. Combining G007-LK/BKM120/Erlotinib significantly
enhanced numbers of cells in G1 phase. Also the differential effect on CCND1,
CMYC, CDKN1A and CDKN1B expression was most prominent with G007-
LK/BKM120/Erlotinib treatment. Protein levels of CyclinD1 and cMYC reflected
transcript expression (Fig. 3D).
In HCT-15 cells, apoptosis was not significantly changed with any
combinations of inhibitor treatments, although G007-LK treatment showed a tendency
towards more apoptosis (Fig. 3E). G007-LK treatment did not affect cell cycle
distribution (Fig. 3F and Supplementary Fig. S14), reflecting the lack of growth
reduction. Erlotinib significantly increased the number of cells in the G1 phase
compared to control when combined with either G007-LK or BKM120, reflected by a
significantly increased CDKN1A and CDKN1B expression (Fig. 3G and
Supplementary Fig. S14). The impact of inhibitors on CMYC and CCND1 expression
was more complex. Both CMYC and CCND1 transcription was increased by both
BKM120 and Erlotinib treatment, while G007-LK reduced their expression both
alone and in combination with BKM120 and/or Erlotinib. The protein levels of
CyclinD1 and cMYC reflected well their transcript levels (Fig. 3H).
p53 is a tumor suppressor known to be involved in many cancer relevant
functions, including growth arrest, apoptosis and colony formation (26), presumably
through interactions with cMYC (27,28). Both COLO320DM and HCT-15 cells carry
TP53 mutations (Supplementary Fig. S1) which compromise their ability to induce
G1 arrest through enhanced CDKN1A expression. Inhibitor treatments, in particular
when combined, significantly increased TP53 expression in COLO320DM
(Supplementary Fig. S14). However, this was not reflected by p53 protein levels
which remained unaltered (Fig. 3D).
In HCT-15 cells, TP53 expression levels were significantly increased by both
single and combined BKM120 and Erlotinib treatment, and attenuated by G007-LK
(Supplementary Fig. S14). TP53 expression levels were well reflected by p53 protein
levels (Fig. 3H).
An esiRNA mediated knock down of TP53 did not change growth response to
inhibitors in COLO3230DM cells. In contrast, HCT-15 cells showed a statistically
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
11
relevant enhancement in their response to combined G007-LK/BKM120 and G007-
LK/Erlotinib treatment with TP53 knock down (Supplementary Fig. S15).
Taken together; In COLO320DM cells combined inhibitor treatments reduced
proliferation and colony formation accompanied by a combination of reduction in
cMYC and CyclinD1 protein levels, enhanced expression of CDKN1A and CDKN1B
cell cycle regulators and induction of apoptosis. In HCT-15 cells the colony forming
ability was reflected by changes in CyclinD1 and cMYC protein levels, while
proliferation was reflected by changes in cMYC, and restrained by upregulated p53.
Inhibitor effects on WNT signaling pathway components
Next, we investigated how inhibitor treatments affected their respective biotargets,
and the WNT/E-catenin and AKT/mTOR signaling pathways. G007-LK exposure
may either stabilize or destabilize its primary biotarget tankyrase (TNKS1/2) in a
contextual way that is currently poorly understood (10,11). In COLO320DM and
HCT-15 protein levels of TNKS1/2 were stabilized upon G007-LK treatment (Fig.
4A-B). Furthermore, AXIN1 and AXIN2, which are structural proteins in the WNT/E￾catenin destruction complex, were stabilized. BKM120 selectively stabilized AXIN1
(although less powerful than G007-LK) in COLO320DM and HCT-15, while it did
not affect AXIN2. Erlotinib did not stabilize either of the two AXIN proteins. AXIN2
stabilization was maximal by single G007-LK treatment, while both BKM120 and
Erlotinib reduced G007-LK mediated AXIN2 stabilization.
Ecatenin is the key transmitter of the canonical WNT signaling pathway, and
its stability is regulated by the destruction complex (29). Active GSK3E targets E￾catenin for ubiquitination and degradation by the proteasome by N-terminal
phosphorylation of E-catenin (at serine 33/37 and threonine 41). GSK3E also
phosphorylates AXIN1/2, enhancing its binding to E-catenin, and further stabilizing
the E-catenin destruction complex (30,31). The activity of GSK3E is regulated by
several pathways, including AKT signaling (32), which inactivates GSK3E through
phosphorylation of serine 9 (pGSK3ES In COLO320DM, both G007-LK and
BKM120 reduced the level of inactive pGSK3ES (Fig. 4C). However, only by
G007-LK treatment, which also stabilized AXIN2, reduction of pGSK3ES led to
increased N-terminal phosphorylation (S33/S37/T41) of E-catenin and reduced levels
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
12
of both total and non-phospho(active)E-catenin (ABC). Reduction of pGSK3ES by
BKM120 (without AXIN2 stabilization) did not increase N-terminal phosphorylation
of E-catenin. This is in line with reports showing that GSK3E activation through AKT
inactivation does not affect canonical WNT signaling (33). Under all conditions, the
levels of pE-catenin(S33/S37/T41) corresponded inversely with the levels of total E￾catenin, indicating that the pE-catenin(S33/S37/T41) was targeted for degradation.
Intriguingly, upon combining G007-LK/BKM120/Erlotinib treatment, N-terminal
phosphorylation of E-catenin was massively increased, accompanied by a strong
reduction of both total, ABC, C-terminal phosphorylated E-catenin and nuclear E￾catenin (Fig. 4C and Supplementary Fig. S16). This effect was seen despite reduced
AXIN1/2 stabilization.
In HCT-15 cells neither G007-LK, BKM120 nor Erlotinib single treatment
affected pGSK3ESlevels (Fig. 4D), although we observed a mild increase of pE￾catenin(S33/S37/T41) levels by G007-LK which, despite massive stabilization of
AXIN1/2, was not reflected by either the total, ABC, C-terminal phosphorylated or
nuclear levels of E-catenin (Fig. 4D and Supplementary Fig. S16). Only when G007-
LK was combined with BKM120 or Erlotinib, levels of pGSK3ES were
moderately reduced, pE-catenin(S33/S37/T41) increased, and the total levels of E￾catenin were reduced. Hence, in HCT-15 cells, BKM120 and Erlotinib acted as
enabling factors for the destabilization of E-catenin by G007-LK. In conclusion,
inhibition of tankyrase by G007-LK alone activated GSK3E and reduced the level of
E-catenin in COLO320DM, while in HCT-15 cells a combination of inhibitors was
needed.
Nuclear E-catenin is a direct regulator of AXIN2 transcription in a negative
feedback loop (34,35). Nuclear levels of E-catenin should therefore positively
correlate to AXIN2 transcription. In both COLO320DM and HCT-15 cells nuclear
levels of ABC correlated with total levels of ABC under all conditions (Fig. 4C-D). In
COLO320DM cells, reduced AXIN2 expression correlated well with G007-LK
mediated reduction of nuclear ABC (Fig. 4C and E). Interestingly, also in HCT-15
cells G007-LK led to a reduction of AXIN2 transcripts, despite absence of altered
nuclear ABC levels (Fig. 4D and F). In HCT-15 cells the levels of ABC were
therefore not predictive for AXIN2 transcription. Previous studies have shown that
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
13
nuclear levels of AXIN2 negatively regulate WNT/E-catenin signaling by binding to
the E-catenin/TCF complex (36).
In COLO320DM nuclear AXIN2 levels correlated positively with total levels
of AXIN2, while negatively with nuclear ABC (Fig. 4 A and C). Hence,
COLO320DM cells exhibited a G007-LK response that is consistent with the classical
understanding of canonical WNT signaling and in accordance with the inhibitory role
of AXIN2 in the nucleus (36). In contrast, in HCT-15 cells nuclear AXIN2 levels
were reduced by G007-LK treatment, despite the overall stabilization of AXIN2 (Fig.
4B and D). It is at current unclear what prevented AXIN2 from entering the nucleus
upon G007-LK treatment in HCT-15 cells. Despite less nuclear AXIN2 in
combination with stable E-catenin levels, which should predict increased CMYC
expression (36), G007-LK reduced transcription of both AXIN2 and CMYC in HCT-
15 cells (Fig. 4F and Supplementary Fig. S14).
Inhibitor effects on AKT/mTOR and EGFR signaling pathway components
Inhibition of PI3K signaling by BKM120 should prevent downstream
activation of AKT (8). In both COLO320DM and HCT-15 cells BKM120
significantly reduced activated AKT (pAKT(S473)), while total AKT levels were
unaffected (Fig 5A and B). Interestingly, in COLO320DM cells, but not in HCT-15
also G007-LK moderately reduced pAKT(S473). Oppositely, Erlotinib enhanced
pAKT(S473) levels in both cell lines. This was unexpected since PI3K signaling is
downstream of EGFR. In both cell lines combined G007-LK/BKM120/Erlotinib
treatment reduced pAKT(S473) levels most effectively.
mTOR is a downstream target of AKT that affects mRNA translation through
regulation of the S6 ribosomal protein (S6RP). In COLO320DM all inhibitors reduced
the active form of S6RP (pS6RP(S240/244)), and the active form of its regulator
P70S6K (pP70S6K(T389)) (Fig. 5C). P70S6K activity was further inhibited by
combined inhibitor treatments, in particular by G007-LK/BKM120/Erlotinib
treatment. Since G007-LK affected both E-catenin and AKT/mTOR signaling in
COLO320DM, the reduced proliferation and colony formation by this inhibitor is
likely to be impacted by the combined alteration of both pathways. In contrast,
BKM120 led to an inhibition of the AKT/mTOR signaling pathway only, while
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
14
Erlotinib enhanced AKT signaling and reduced mTOR activity. Inhibitor
combinations reliably reduced both pathways.
In HCT-15 cells, BKM120 and Erlotinib reduced active P70S6K and S6RP
(Fig. 5D). When combined their effect on pS6RP(S240/244) was highly potentiated,
and reflected the significant reduction in cell proliferation and colony formation.
G007-LK treatment, although clearly stabilizing AXIN1/2, did not affect the levels of
either protein. Hence, tankyrase inhibition appeared to be uncoupled from impacting
both the AKT/mTOR and WNT/β-catenin signaling in HCT-15 cells. Despite
enhanced AKT activation with Erlotinib in both cell lines, both active P70S6K and
S6RP were reduced. This suggests that AKT and mTOR were differentially regulated
with Erlotinib treatment as compared to G007-LK and BKM120 treatment, where
both AKT and P70S6K/S6RP activity were equally regulated.
Erlotinib affects EGFR activity by preventing receptor dimerization, reliably
affecting phosphorylation of tyrosine 1068 (pEGFR(1068)) (37). Due to low
expression of EGFR in COLO320DM (38,39) only weak bands of pEGFR(1068)
were obtained even with high protein concentrations (Fig. 5E, lower panel). As
expected, pEGFR(1068) was reduced with Erlotinib, while it was unaffected by
G007-LK, and enhanced by BKM120. In HCT-15 cells pEGFR(1068) was strongly
reduced by Erlotinib treatment, while both G007-LK and BKM120 induced
pEGFR(1068) protein level (Fig. 5F), which was further increased when the two
were combined. Increased pEGFR(1068) was efficiently attenuated by Erlotinib
treatment. In conclusion, combined G007-LK/BKM120/Erlotinib treatment reduced
expression of hallmark proteins of both E-catenin and AKT/mTOR signaling
pathways in both COLO320DM and HCT-15 cells, and EGFR activity in HCT-15
cells.
GSK3β is a protein functionally linking AKT and WNT signaling and has
been implied in the tankyrase/AXIN effect on RAS signaling (32,40). Since we
observed that tankyrase inhibition affected both WNT/E-catenin and AKT/mTOR
signaling in COLO320DM but not in HCT-15 cells, we tested whether the inhibitor
impact on RAS expression could be indicative for the differential response of both
cell lines. Activated GSK3β has been shown to destabilize both β-catenin and RAS in
CRC cells (40), and RAS is an upstream regulator of AKT (41). However, G007-LK
at the tested dose did not affect RAS protein level neither in COLO320DM nor in
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
15
HCT-15 cells (Fig. 5G-H). There were also no detectable changes in RAS levels upon
BKM120 or Erlotinib treatment in KRAS wild type COLO320DM cells, while in
KRAS mutant HCT-15 cells BKM120 slightly and Erlotinib substantially reduced
RAS levels in all treatment combinations, and most substantially upon G007-
LK/BKM120/Erlotinib treatment.
Inhibitor impacts on tankyrase/E-catenin complex formation
AXIN1/2 together with GSK3E APC and CK1α are components of the E￾catenin destruction complex (42,43). Since E-catenin was reduced by tankyrase
inhibition (G007-LK) in COLO320DM, but not in HCT-15 cells, we investigated by
immunofluorescence whether tankyrase and E-catenin co-localized upon inhibitor
treatments in both cell lines.
In COLO320DM, E-catenin was localized predominantly cytoplasmic and
nuclear with control treatment. Following G007-LK treatment, E-catenin localization
was reduced in the nucleus and co-localized with tankyrase containing puncta in the
cytoplasm (Fig. 6A, Supplementary Fig. S17). Co-localization of tankyrase and E￾catenin was observed in all treatments containing G007-LK despite attenuated
AXIN1/2 levels by BKM120 and Erlotinib treatment (Fig. 4A). No tankyrase puncta
were observed upon BKM120 and/or Erlotinib treatment.
In HCT-15 cells, both tankyrase and E-catenin were predominantly found at
the cell membrane upon control treatment. G007-LK induced tankyrase containing
puncta, but E-catenin was not co-localized in these puncta (Fig. 6B, Supplementary
Fig. S17). Although G007-LK induced N-terminal E-catenin phosphorylation and
reduced E-catenin levels in combination with BKM120 or Erlotinib, in HCT-15 cells,
E-catenin degradation appears to be independent from a complex containing
tankyrase.
Combined inhibitor treatment exerts additive or synergistic effects on CRC
xenograft growth reduction
To explore the possible additive effects of combined inhibitor treatments
against tumor growth in vivo, xenografts were established using COLO320DM-Luc2,
HCT-15-Luc2 and HCT-15 cells in CB17SCID mice. Mice carrying the xenografts
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
16
were treated with moderate doses of the compounds (G007-LK [10 mg/kg], BKM120
[3 mg/kg] and Erlotinib [15 mg/kg], n=10).
In COLO320DM, single G007-LK or BKM120 treatment resulted in 17% and
6% tumor size reduction, respectively, while Erlotinib led to a 46% reduction.
Combined G007-LK/BKM120 displayed an augmented effect (49% reduction).
Treatment with Erlotinib in combination with G007-LK or BKM120 resulted in 39%
or 25% tumor reduction, respectively, a change without statistical relevance compared
to a combined G007-LK/BKM120/Erlotinib treatment (39% reduction, Fig. 7A,
Supplementary Fig. S16 and S17). Treatment at elevated drug doses (G007-LK [50
mg/kg], BKM120 [15 mg/kg] and Erlotinib [30 mg/kg]) produced a more significant
tumor size reduction. At these doses, single agent treatment with G007-LK, BKM120
and Erlotinib resulted in 78%, 75% and 55% tumor size reduction, respectively, while
combined G007-LK/BKM120 treatment enhanced the effect of single treatments to an
86% reduction, and more pronounced upon G007-LK/BKM120/Erlotinib treatment
(91% reduction, Fig. 7A, Supplementary Fig. S18 and S19). Molecular analysis of
tumor protein extracts reflects in vitro analysis (Supplementary Fig. S20).
In HCT-15 xenografts, single agent treatments caused only trends at moderate
doses compared to control (G007-LK; 3%, BKM120; 18%, Erlotinib; 21%). Only
when Erlotinib was combined with a G007-LK/BKM120 treatment a more robust
reduction on tumor growth was observed (34% against 16% reduction; Fig. 7B and
Supplementary Fig. S18 and S19). At elevated drug doses single treatment with
G007-LK, BKM120 and Erlotinib resulted in 3%, 76% and 8% tumor size reduction,
respectively. Combined G007-LK/BKM120/Erlotinib treatment resulted in a striking
94% tumor size reduction compared to the control, effectively halting tumor growth
(Fig. 7B, Supplementary Fig. S18 and S19).
Mouse body weights (Supplementary Fig. S21) indicated overall tolerability of
the treatment regimes, including to the increased dose G007-LK/BKM120/Erlotinib
treatment.

Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
17
DISCUSSION
In this study we have investigated a combinatorial therapeutic approach
combining tankyrase/WNT, PI3K and EGFR inhibition to explore treatment strategies
for CRC. While both PI3K and EGFR inhibitors are in a clinical stage (6-9), tankyrase
inhibitors are still at a preclinical exploratory stage due to a limited efficacy of WNT
inhibitors in current preclinical CRC models (11,12) and to intestinal cytotoxicity
attributed to WNT pathway inhibition (12,14). In addition, the contextual background
of CRC cells required to render WNT inhibitors efficacious remains poorly
understood, although the length of the APC protein is proposed to play an important
role (44).
Our study revealed that G007-LK exerts a dual effect on E-catenin and
AKT/mTOR activity in the tankyrase inhibitor sensitive (affecting proliferation)
COLO320DM and SW403 cell lines, but not in the inhibitor insensitive HCT-15 and
SW480 cells. In HCT-15 cells, G007-LK rather initiated EGFR feedback activation.
Contextual differences such as the mutational background of cells may be important
for enabling or blocking this dual pathway response.
In both COLO320DM and HCT-15 cells Erlotinib enhanced activation of
AKT, while reducing the activity of both EGFR and mTOR effectors. Similar effects
have been observed with EGFR inhibition in non-small cell lung cancer where AKT
activation was induced either through activation of STAT3 (45), or ERK-2 (46).
Enhanced AKT activity could also be induced by reduced S6RP (mTOR) activity
which may relieve the TORC1 feedback regulation and enhance the interaction
between PI3K and IRS proteins (47).
In KRAS mutant HCT-15 cells, G007-LK neither affected the levels of ABC
nor proliferation, despite the classical stabilization of both tankyrase and AXIN1/2. A
possible explanation may be the sub-cellular distributions of E-catenin and tankyrase.
In Drosophila, active RAS signaling counteracts AXIN mediated degradation of
Armadillo (the Drosophila E-catenin) by recruiting the destruction complex to the cell
membrane (48). In accordance, dual inhibition of tankyrase and MEK has shown to be
efficacious in KRAS mutant CRCs (12,17). We show here that with moderate inhibitor
doses, combined tankyrase/PI3K/EGFR inhibition more efficiently reduced cell
proliferation in vitro than combined tankyrase/MEK inhibition in some CRCs,
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
18
including KRAS wild type COLO320DM. Hence, combined tankyrase/PI3K/EGFR
inhibition may be a potent treatment option that is independent of a KRAS mutant
background.
While both KRAS mutations and the degree of APC truncation could affect the
sensitivity to tankyrase inhibition (12,44), we observed that combined
BKM120/Erlotinib treatment sensitized HCT-15 cells to G007-LK treatment, reduced
ABC, proliferation and the colony forming ability. Possibly because Erlotinib
attenuated the induced EGFR feedback activation mediated by G007-LK and
BKM120, accompanied by a reduction in RAS protein levels. Combined G007-
LK/BKM120/Erlotinib therefore most potently reduced both WNT, AKT, mTOR,
EGFR and RAS signaling in both COLO320DM and HCT-15 cells.
Tankyrase inhibition has shown to induce puncta in CRC cells containing
functional components of the β-catenin destruction complex, including tankyrase, β-
catenin and AXIN1/2 (43). In our study, tankyrase inhibition alone induced
stabilization of both AXIN1 and AXIN2. However, when combined with BKM120
and Erlotinib, AXIN1/2 stabilization was reduced to the level of control with G007-
LK/BKM120/Erlotinib treatment. Despite reduced AXIN1/2 stabilization, tankyrase
containing puncta were formed in the presence of G007-LK in all treatment regimes.
β-catenin co-localized with these puncta only in COLO320DM but not in HCT-15
cells. Notwithstanding, G007-LK/BKM120/Erlotinib treatment strongly reduced ABC
in both cell lines. We therefore assume that AXIN1/2 stabilization is not required for
tankyrase puncta formation, and neither correlates with degradation of β-catenin.
Our study shows that combined tankyrase/PI3K/EGFR inhibition, mutually
affecting in a context dependent manner components of the WNT/E-catenin,
AKT/mTOR, EGFR and RAS signaling pathways, may provide a treatment strategy
for CRC cells with divergent mutational backgrounds. The tankyrase/PI3K/EGFR
inhibitor combination also led to a significant reduction of tumor size in vivo without
affecting animal body weight. Although possible adverse effect of this combination
will have to be further explored, tankyrase/PI3K/EGFR inhibition may provide a
versatile combination expanding the possibilities of tankyrase/MEK inhibition in
KRAS mutant CRC tumors.

Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
19
ACKNOWLEDGMENT
We would like to acknowledge René Beyer and the IOI Oleo GmbH for the Mygliol
810N drug carrier solution, and Jennifer Dembinski for technical support during
initial xenograft experiments.
Reference List
1. Segditsas S, Tomlinson I. Colorectal cancer and genetic alterations in the Wnt pathway.
Oncogene 2006;25:7531-7
2. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996;87:159-70
3. Fearon ER. Molecular genetics of colorectal cancer. Annual review of pathology 2011;6:479-
507
4. Markman B, Tabernero J, Krop I, Shapiro GI, Siu L, Chen LC, et al. Phase I safety,
pharmacokinetic, and pharmacodynamic study of the oral phosphatidylinositol-3-kinase and
mTOR inhibitor BGT226 in patients with advanced solid tumors. Annals of oncology :
official journal of the European Society for Medical Oncology 2012;23:2399-408
5. Brown KK, Toker A. The phosphoinositide 3-kinase pathway and therapy resistance in
cancer. F1000prime reports 2015;7:13
6. Pollack VA, Savage DM, Baker DA, Tsaparikos KE, Sloan DE, Moyer JD, et al. Inhibition of
epidermal growth factor receptor-associated tyrosine phosphorylation in human carcinomas
with CP-358,774: dynamics of receptor inhibition in situ and antitumor effects in athymic
mice. The Journal of pharmacology and experimental therapeutics 1999;291:739-48
7. Wakeling AE, Barker AJ, Davies DH, Brown DS, Green LR, Cartlidge SA, et al. Specific
inhibition of epidermal growth factor receptor tyrosine kinase by 4-anilinoquinazolines. Breast
cancer research and treatment 1996;38:67-73
8. Maira SM, Pecchi S, Huang A, Burger M, Knapp M, Sterker D, et al. Identification and
characterization of NVP-BKM120, an orally available pan-class I PI3-kinase inhibitor.
MolCancer Ther 2012;11:317-28
9. Tzeng HE, Yang L, Chen K, Wang Y, Liu YR, Pan SL, et al. The pan-PI3K inhibitor GDC-
0941 activates canonical WNT signaling to confer resistance in TNBC cells: resistance
reversal with WNT inhibitor. Oncotarget 2015;6:11061-73
10. Huang SM, Mishina YM, Liu S, Cheung A, Stegmeier F, Michaud GA, et al. Tankyrase
inhibition stabilizes axin and antagonizes Wnt signalling. Nature 2009;461:614-20
11. Waaler J, Machon O, Tumova L, Dinh H, Korinek V, Wilson SR, et al. A novel tankyrase
inhibitor decreases canonical Wnt signaling in colon carcinoma cells and reduces tumor
growth in conditional APC mutant mice. Cancer Res 2012;72:2822-32
12. Lau T, Chan E, Callow M, Waaler J, Boggs J, Blake RA, et al. A novel tankyrase small￾molecule inhibitor suppresses APC mutation-driven colorectal tumor growth. Cancer Res
2013;73:3132-44
13. Zhong L, Ding Y, Bandyopadhyay G, Waaler J, Borgeson E, Smith S, et al. The PARsylation
activity of tankyrase in adipose tissue modulates systemic glucose metabolism in mice.
Diabetologia 2016;59:582-91
14. Zhong Y, Katavolos P, Nguyen T, Lau T, Boggs J, Sambrone A, et al. Tankyrase Inhibition
Causes Reversible Intestinal Toxicity in Mice with a Therapeutic Index < 1. Toxicologic
pathology 2016;44:267-78
15. Casas-Selves M, Kim J, Zhang Z, Helfrich BA, Gao D, Porter CC, et al. Tankyrase and the
canonical Wnt pathway protect lung cancer cells from EGFR inhibition. Cancer Res
2012;72:4154-64
16. Tenbaum SP, Ordonez-Moran P, Puig I, Chicote I, Arques O, Landolfi S, et al. beta-catenin
confers resistance to PI3K and AKT inhibitors and subverts FOXO3a to promote metastasis in
colon cancer. NatMed 2012;18:892-901
17. Schoumacher M, Hurov KE, Lehar J, Yan-Neale Y, Mishina Y, Sonkin D, et al. Inhibiting
Tankyrases sensitizes KRAS-mutant cancer cells to MEK inhibitors via FGFR2 feedback
signaling. Cancer research 2014;74:3294-305
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
20
18. Voronkov A, Holsworth DD, Waaler J, Wilson SR, Ekblad B, Perdreau-Dahl H, et al.
Structural basis and SAR for G007-LK, a lead stage 1,2,4-triazole based specific tankyrase 1/2
inhibitor. JMedChem 2013;56:3012-23
19. Schepers AG, Snippert HJ, Stange DE, van den Born M, van Es JH, van de Wetering M, et al.
Lineage tracing reveals Lgr5+ stem cell activity in mouse intestinal adenomas. Science
2012;337:730-5
20. Tetsu O, McCormick F. Beta-catenin regulates expression of cyclin D1 in colon carcinoma
cells. Nature 1999;398:422-6
21. Lee T, Yao G, Nevins J, You L. Sensing and integration of Erk and PI3K signals by Myc.
PLoS computational biology 2008;4:e1000013
22. Bretones G, Delgado MD, Leon J. Myc and cell cycle control. Biochimica et biophysica acta
2015;1849:506-16
23. Perez-Roger I, Kim SH, Griffiths B, Sewing A, Land H. Cyclins D1 and D2 mediate myc￾induced proliferation via sequestration of p27(Kip1) and p21(Cip1). EMBO J 1999;18:5310-
20
24. Jeanteur P. [The role of APC in colonic cancerogenesis: zeroing in on Myc]. Bulletin du
cancer 1998;85:925-8
25. Gartel AL, Radhakrishnan SK. Lost in transcription: p21 repression, mechanisms, and
consequences. Cancer Res 2005;65:3980-5
26. Nair HK, Rao KV, Aalinkeel R, Mahajan S, Chawda R, Schwartz SA. Inhibition of prostate
cancer cell colony formation by the flavonoid quercetin correlates with modulation of specific
regulatory genes. Clinical and diagnostic laboratory immunology 2004;11:63-9
27. Vafa O, Wade M, Kern S, Beeche M, Pandita TK, Hampton GM, et al. c-Myc can induce
DNA damage, increase reactive oxygen species, and mitigate p53 function: a mechanism for
oncogene-induced genetic instability. Molecular cell 2002;9:1031-44
28. Abraham SA, Hopcroft LE, Carrick E, Drotar ME, Dunn K, Williamson AJ, et al. Dual
targeting of p53 and c-MYC selectively eliminates leukaemic stem cells. Nature
2016;534:341-6
29. Stamos JL, Weis WI. The beta-catenin destruction complex. Cold Spring Harbor perspectives
in biology 2013;5:a007898
30. Yamamoto H, Kishida S, Kishida M, Ikeda S, Takada S, Kikuchi A. Phosphorylation of axin,
a Wnt signal negative regulator, by glycogen synthase kinase-3beta regulates its stability. The
Journal of biological chemistry 1999;274:10681-4
31. Kim SE, Huang H, Zhao M, Zhang X, Zhang A, Semonov MV, et al. Wnt stabilization of
beta-catenin reveals principles for morphogen receptor-scaffold assemblies. Science
2013;340:867-70
32. McCubrey JA, Steelman LS, Bertrand FE, Davis NM, Sokolosky M, Abrams SL, et al. GSK-3
as potential target for therapeutic intervention in cancer. Oncotarget 2014;5:2881-911
33. Ng SS, Mahmoudi T, Danenberg E, Bejaoui I, de Lau W, Korswagen HC, et al.
Phosphatidylinositol 3-kinase signaling does not activate the wnt cascade. The Journal of
biological chemistry 2009;284:35308-13
34. Lustig B, Jerchow B, Sachs M, Weiler S, Pietsch T, Karsten U, et al. Negative feedback loop
of Wnt signaling through upregulation of conductin/axin2 in colorectal and liver tumors.
Molecular and cellular biology 2002;22:1184-93
35. Jho EH, Zhang T, Domon C, Joo CK, Freund JN, Costantini F. Wnt/beta-catenin/Tcf signaling
induces the transcription of Axin2, a negative regulator of the signaling pathway. MolCell
Biol 2002;22:1172-83
36. Rennoll SA, Konsavage WM, Jr., Yochum GS. Nuclear AXIN2 represses MYC gene
expression. Biochemical and biophysical research communications 2014;443:217-22
37. Sette G, Salvati V, Mottolese M, Visca P, Gallo E, Fecchi K, et al. Tyr1068-phosphorylated
epidermal growth factor receptor (EGFR) predicts cancer stem cell targeting by erlotinib in
preclinical models of wild-type EGFR lung cancer. Cell death & disease 2015;6:e1850
38. Dunn EF, Iida M, Myers RA, Campbell DA, Hintz KA, Armstrong EA, et al. Dasatinib
sensitizes KRAS mutant colorectal tumors to cetuximab. Oncogene 2011;30:561-74
39. Koizumi F, Kanzawa F, Ueda Y, Koh Y, Tsukiyama S, Taguchi F, et al. Synergistic
interaction between the EGFR tyrosine kinase inhibitor gefitinib (“Iressa”) and the DNA
topoisomerase I inhibitor CPT-11 (irinotecan) in human colorectal cancer cells. International
journal of cancer Journal international du cancer 2004;108:464-72
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
21
40. Cha PH, Cho YH, Lee SK, Lee J, Jeong WJ, Moon BS, et al. Small-molecule binding of the
axin RGS domain promotes beta-catenin and Ras degradation. Nature chemical biology
2016;12:593-600
41. Campbell M, Allen WE, Sawyer C, Vanhaesebroeck B, Trimble ER. Glucose-potentiated
chemotaxis in human vascular smooth muscle is dependent on cross-talk between the PI3K
and MAPK signaling pathways. Circulation research 2004;95:380-8
42. Martino-Echarri E, Brocardo MG, Mills KM, Henderson BR. Tankyrase Inhibitors Stimulate
the Ability of Tankyrases to Bind Axin and Drive Assembly of beta-Catenin Degradation￾Competent Axin Puncta. PloS one 2016;11:e0150484
43. Thorvaldsen TE, Pedersen NM, Wenzel EM, Schultz SW, Brech A, Liestol K, et al. Structure,
Dynamics and Functionality of Tankyrase Inhibitor-induced Degradasomes. Molecular cancer
research : MCR 2015
44. Tanaka N, Mashima T, Mizutani A, Sato A, Aoyama A, Gong B, et al. APC mutations as a
potential biomarker for sensitivity to tankyrase inhibitors in colorectal cancer. Molecular
cancer therapeutics 2017
45. Wu K, Chang Q, Lu Y, Qiu P, Chen B, Thakur C, et al. Gefitinib resistance resulted from
STAT3-mediated Akt activation in lung cancer cells. Oncotarget 2013;4:2430-8
46. Toulany M, Minjgee M, Saki M, Holler M, Meier F, Eicheler W, et al. ERK2-dependent
reactivation of Akt mediates the limited response of tumor cells with constitutive K-RAS
activity to PI3K inhibition. Cancer biology & therapy 2014;15:317-28
47. Ebi H, Corcoran RB, Singh A, Chen Z, Song Y, Lifshits E, et al. Receptor tyrosine kinases
exert dominant control over PI3K signaling in human KRAS mutant colorectal cancers. The
Journal of clinical investigation 2011;121:4311-21
48. Hall ET, Verheyen EM. Ras-activated Dsor1 promotes Wnt signaling in Drosophila
development. Journal of cell science 2015;128:4499-511

FIGURE LEGENDS
Fig. 1. Tankyrase inhibition (G007-LK) enhanced the growth inhibitory effects
of both PI3K/AKT (BKM120) and EGFR (Erlotinib) inhibition in human
colorectal cancer cell lines in vitro. Combined tankyrase/PI3K/EGFR inhibition
was more effective than combined tankyrase/MEK inhibition
Cell confluence at experimental endpoint of the human colorectal cancer cell lines
COLO320DM (A and C; 14 days incubation) and HCT-15 (B and D; 10 days
incubation) in the presence of pathway inhibitors as indicated. (C and D) demonstrate
advantageous treatment with combined tankyrase/PI3K/EGFR inhibition on cell
growth compared to combined tankyrase/MEK inhibition at selected, moderate
inhibitor doses. The figure show representative graphs from at least 3 separate
biological experiments, each with at least 3 technical replicates. * p<0.05 by two￾tailed t-test. Complete statistical analysis is found in Supplementary Statistics.
Fig. 2. Combined inhibitor treatments reduced the colony forming ability of
colorectal cancer cells and induced molecular characteristics of differentiation
The ability to form colonies upon treatment with inhibitors as indicated reflected
proliferation in COLO320DM cells (A), but not in HCT-15 (B) cells. Data represent
mean relative colony number of 3 (A) and 2 (B) biological experiments, respectively,
each with 6 technical replicates. Relative expression levels of LGR5 and KRT20 was
analyzed after incubation with inhibitors as indicated in COLO320DM (C) and HCT￾Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
22
15 (D) cells. Gene expression is normalized to internal GAPDH levels. Data represent
mean relative expression values (+/- STDEV) of 3 independent experiments, each
with 3 technical replicates. * denotes significance (p<0.05 by two-tailed t-test)
compared to DMSO control, unless otherwise is depicted. Complete statistical
analysis is found in Supplementary Statistics.
Fig. 3. Inhibitor mediated changes in proliferation and colony forming ability
were reflected by changes in cell cycle distribution and apoptosis through an
altered balance between cell cycle genes
Fold induction of apoptotic (Annexin V+) cells after incubation with inhibitors as
indicated in COLO320DM (A) and HCT-15 (E) cells compared to DMSO control
treatment. Changes in percentage of cells in G1 phase of the cell cycle upon inhibitor
treatments in COLO320DM (B) and HCT-15 (F) cells. Expression levels of CCND1,
and CDKN1A upon incubation with inhibitors as indicated in COLO320DM (C) and
HCT-15 (G) cells. Gene expression is normalized to internal GAPDH levels. Data
represent mean relative expression values (+/- STDEV) of 3 independent experiments,
each with 3 technical replicates. Representative regulations in CyclinD1, cMYC and
p53 protein levels are shown in COLO320DM (D) and HCT-15 (H). * p<0.05 by two￾tailed t-test, ** p<0.05 by one-tailed t-test. Complete statistical analysis is found in
Supplementary Statistics.
Fig. 4. Effect of G007-LK, BKM120 and Erlotinib treatment on WNT signaling
pathway components
Regulations of TNKS1/2, AXIN1, AXIN2 (A; COLO320DM, B; HCT-15), total and
inactive GSK3β (phospho serine 9 GSK3β; P-GSK3β), N-terminal phosphorylated β-
catenin (phospho serine 33/37/threonine 41 β-catenin; P-β-catenin), total β-catenin
and non-phospho (active) β-catenin (ABC) protein levels in total cell extracts (C;
COLO320DM and D; HCT-15, upper panel), and nuclear protein levels of AXIN2
and ABC proteins (C and D, lower panel) upon incubation with inhibitors as
indicated. Relative AXIN2 expression levels normalized to internal GAPDH level are
indicated in (E; COLO320DM) and (F; HCT-15). Data represent mean relative
expression values (+/- STDEV) of 3 independent experiments, each with 3 technical
replicates. * p<0.05 by two-tailed t-test. Complete statistical analysis is found in
Supplementary Statistics.
Fig. 5. Effect of G007-LK, BKM120 and Erlotinib treatment on AKT/mTOR and
EGFR signaling pathway components
Protein levels of total and active AKT (phospho serine 473 AKT; P-AKT) in
COLO320DM (A) and HCT-15 (B) cells. Protein levels of total and active P70S6K
(phospho tyrosine389-P70S6K; P-P70S6K), and total and active S6RP (phospho
serine240/244-S6RP; P-S6RP) in COLO320DM (C) and HCT-15 (D) cells. Protein
levels of total and active EGFR (phospho tyrosine 1068 EGFR; P-EGFR) protein
levels in COLO320DM (E) and HCT-15 (F) cells, and total RAS in COLO320DM
(G) and HCT-15 (H) cells. 15-20 μg protein was probed with the indicated antibodies,
and 40 μg for P-EGFR in the lower panel in (E). Data represent one experiment of at
least two.
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-7786.MCR-17-0362
23
Fig. 6. Effect of G007-LK, BKM120 and Erlotinib treatment on tankyrase/β-
catenin complex formation
Representative images of immunofluorescent staining of β-catenin (left), TNKS1/2
(middle) and merge (right) of β-catenin (green), TNKS1/2 (red) and DAPI (blue) in
COLO320DM (A) and HCT-15 (B) cells treated with inhibitors as indicated. A co￾localization mask displaying TNKS/β-catenin co-localization in white was
superimposed on the merged images. Arrows indicate β-catenin puncta (left image),
TNKS1/2 puncta (middle image) and TNKS/β-catenin co-localization (right image).
Scale bar; 10 μm.
Fig. 7. Combined treatments with G007-LK, BKM120 and Erlotinib resulted in
additive or synergistic effects against colorectal cancer xenograft growth in vivo
Dot plots showing tumor end weights relative to control (mean set to 0%) for
COLO320DM-Luc2 (A), and HCT-15-Luc2 (low dose) and HCT-15 (high dose)
xenograft experiments (B). Treatment doses used are shown in the x-axis. Each dot
shows single tumors and black bold lines (-) depict mean values. Synergistic (blue),
additive (green) and reductive (red) effects upon combinatorial treatments are
highlighted and shown as numerical values. Selected statistically significant
differences are highlighted: Rank sum tests: * P <0.05, t-tests: ** P <0.05. n =10 mice
per group.
Author manuscripts have been peer rviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst 1-
Fig. 1. Tankyrase inhibition (G007-LK) enhanced the
growth inhibitory effects of both PI3K/AKT (BKM120)
and EGFR (Erlotinib) inhibition in human colorectal
cancer cell lines in vitro. Combined tankyrase/PI3K/EG￾FR inhibition was more effective than combined
tankyrase/MEK inhibition
Cell confluence at experimental endpoint of the human
colorectal cancer cell lines COLO320DM (A and C; 14 days
incubation) and HCT-15 (B and D; 10 days incubation) in
the presence of pathway inhibitors as indicated. (C and D)
demonstrate advantageous treatment with combined
tankyrase/PI3K/EGFR inhibition on cell growth compared to
combined tankyrase/MEK inhibition at selected, moderate
inhibitor doses. The figure show representative graphs from
at least 3 separate biological experiments, each with at
least 3 technical replicates. * p<0.05 by two-tailed t-test.
Complete statistical analysis is found in Supplementary
Statistics.
Fig. 3. Inhibitor mediated changes in proliferation and colony forming ability were reflected by
changes in cell cycle distribution and apoptosis through an altered balance between cell cycle
genes
Fold induction of apoptotic (Annexin V+) cells after incubation with inhibitors as indicated in COLO320DM
(A) and HCT-15 (E) cells compared to DMSO control treatment. Changes in percentage of cells in G1
phase of the cell cycle upon inhibitor treatments in COLO320DM (B) and HCT-15 (F) cells. Expression
levels of CCND1, and CDKN1A upon incubation with inhibitors as indicated in COLO320DM (C) and
HCT-15 (G) cells. Gene expression is normalized to internal GAPDH levels. Data represent mean
relative expression values (+/- STDEV) of 3 independent experiments, each with 3 technical replicates.
Representative regulations in CyclinD1, cMYC and p53 protein levels are shown in COLO320DM (D)
Fig. 5. Effect of G007-LK, BKM120 and Erlotinib treatment on
AKT/mTOR and EGFR signaling pathway components
Protein levels of total and active AKT (phospho serine 473 AKT;
P-AKT) in COLO320DM (A) and HCT-15 (B) cells. Protein levels of
total and active P70S6K (phospho tyrosine389-P70S6K; P-P70S6K),
and total and active S6RP (phospho serine240/244-S6RP; P-S6RP)
in COLO320DM (C) and HCT-15 (D) cells. Protein levels of total and
active EGFR (phospho tyrosine 1068 EGFR; P-EGFR) protein levels
in COLO320DM (E) and HCT-15 (F) cells, and total RAS in
COLO320DM (G) and HCT-15 (H) cells. 15-20 μg protein was
probed with the indicated antibodies, and 40 μg for P-EGFR in the
lower panel in (E). Data represent one experiment of at least two.
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst on December 8, 2017; DOI: 10.1158/1541-
E-catenin TNKS merge
Downloaded from mcr.aacrjournals.org on December 10, 2017. © 2017 American Association for Cancer Research.
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author Manuscript Published OnlineFirst
Fig. 7. Combined treatments with G007-LK, BKM120 and Erlotinib resulted in additive or synergistic
effects against colorectal cancer xenograft growth in vivo
Dot plots showing tumor end weights relative to control (mean set to 0%) for COLO320DM-Luc2 (A), and
HCT-15-Luc2 (low dose) and HCT-15 (high dose) xenograft experiments (B). Treatment doses used are shown in
the x-axis. Each dot shows G007-LK single tumors and black bold lines (-) depict mean values. Synergistic (blue), additive
(green) and reductive (red) effects upon combinatorial treatments are highlighted and shown as numerical values.
Selected statistically significant differences are highlighted: Rank sum tests