cardiac glycosides, a novel treatment for neuroblastoma ... · pdf fileii cardiac glycosides,...
TRANSCRIPT
80 i
Cardiac Glycosides, a Novel Treatment for Neuroblastoma: Efficacy and Mechanism
by
Paulo Clasio De Gouveia
A thesis submitted in conformity with the requirements for the degree of Master of Science
Institute of Medical Science University of Toronto
© Copyright by Paulo De Gouveia 2010
ii
Cardiac Glycosides, a Novel Treatment for Neuroblastoma: Efficacy and Mechanism
Paulo De Gouveia
Master’s of Science Institute of Medical Science
University of Toronto 2010
Abstract
In an attempt to identify agents that specifically target neuroblastoma (NB) tumour-initiating
cells (TIC) we performed drug screens using libraries of bioactive compounds. Cardiac
glycosides (CGs) were the largest class of drugs identified with antitumour activity. At high CG
doses inhibitory effects on the Na+/K+-ATPase induce cardiotoxicity; therefore, CG analogues
were designed in an attempt to separate the effects on NB cells from cardiotoxicity. We
identified RIDK34 as our lead compound from a structure-activity-relationship analysis (IC50 8
nM). RIDK34 contains a unique oxime group and shows increasing potency against NB TICs.
The Na+/K+-ATPase is a target for the apoptotic activity of digoxin and RIDK34, whereby a
signaling cascade involving Src and ERK may induce apoptosis. Furthermore, we predict that
signaling activation does not require inactivation of the Na+/K+-ATPase and subsequent
deregulation of [Na+]i and [K+]I gradients. Thus CGs and particularly RIDK34 may be expected
to display diminished cardiotoxicity and greater therapeutic potential.
iii
Acknowledgments
I would like to thank everyone who has made this work possible, beginning with my family and
friends who have supported my educational pursuits. Both my mother and brother have inspired
me to work hard and to pursue what I enjoy most. They have also taught me invaluable lessons
that I will carry on for the rest of my life. I would also like to thank my extended family; namely,
my cousins who are the equivalent of brothers and sisters. They have reminded me not only to be
dedicated to my work but also to enjoy what it is I get to do.
The members of the Irwin lab deserve credit for supporting this work, including Dan, Lynn,
Fiona, Jacky, Jennifer, Joanne, Ian, Alvaro and Paola. At one point or another, I’ve come to each
lab member with questions and their willingness to help me has been extraordinary. Also
members of the Kaplan/Miller lab, particularly those in the neuroblastoma group have been
instrumental in making this work possible. Of note, I would like to thank Kristen Smith and
Mayumi Fujitani for their work that contributed to this project on cardiac glycosides. From
working alongside members of both these lab groups, I’ve formed wonderful bonds and have
acquired skills as a scientist that will stay with me to the next phase of my career.
I would also like to thank some of the people whom I’ve collaborated with including Jaimie and
Myl from Dr. Clifford Lingwood’s laboratory. Both Jamie and Myl were instrumental in making
the novel cardiac glycoside compounds and for generating interesting discussions about these
compounds. Also, I would like to thank Clifford Lingwood and Aaron Schimmer- my two
committee members for providing positive feedback and different perspectives within and
outside of committee meetings. Lastly, I would like to thank David Kaplan and Meredith Irwin
for allowing me to pursue a master’s in their respective labs. Individually, David has motivated
me to think about the big picture and has encouraged me to be a well-rounded scientist. Also, as
a supervisor, Meredith has been incredibly supportive over the past two years and I feel very
fortunate to have had the opportunity to train under her guidance. Thank you everyone!
iv
Table of Contents
Acknowledgments ........................................................................................................................ iii
List of Tables ............................................................................................................................... vii
List of Figures............................................................................................................................. viii
List of Abbreviations .................................................................................................................... x
Chapter 1 Review of Literature................................................................................................... 1 1.1 Overview of Neuroblastoma ...............................................................................................................1 1.1.1 Introduction .....................................................................................................................................1 1.1.2 NB treatments, staging and prognosis .............................................................................................1 1.1.3 Genetic Predisposition to NB ..........................................................................................................2 1.1.4 Genetic Characteristics of NB .........................................................................................................3 1.1.5 NB Characteristics ...........................................................................................................................5 1.1.6 Neuroblastoma Tumour-Initiating Cells..........................................................................................5 1.2 Sodium Potassium ATPase .................................................................................................................8 1.2.1 Introduction .....................................................................................................................................8 1.2.2 Structure and Function of the Na+/K+-ATPase................................................................................9 1.2.3 Subunit Diversity and Expression of the Na+/K+-ATPase.............................................................10 1.2.4 Regulation of the Na+/K+-ATPase .................................................................................................12 1.2.5 Na+/K+-ATPase: a novel anticancer target ....................................................................................13 1.3 Cardiac Glycosides............................................................................................................................15 1.3.1 Introduction ...................................................................................................................................15 1.3.2 Chemical structure .........................................................................................................................15 1.3.3 CG binding site on the Na+/K+-ATPase ........................................................................................16 1.3.4 CGs as inotropic and cardiotoxic agents .......................................................................................17 1.3.5 The role of the Na+/K+-ATPase and CG antitumour activity ........................................................20 1.3.6 CGs: Mechanism of antitumour activity .......................................................................................23 1.3.7 Cardiac Glycoside: epidemiological evidence of anticancer properties........................................25 1.3.8 CGs: anti-tumour activity in vivo and in clinical studies...............................................................25 1.3.9 Novel Strategies to circumvent Cardiac Glycoside induced toxicity ............................................27
Chapter 2 Rationale, Aims and Hypothesis.............................................................................. 28
v
2.1 Rationale............................................................................................................................................28 2.2 Hypothesis .........................................................................................................................................29 2.3 Project Aims ......................................................................................................................................29
Chapter 3 Material and Methods .............................................................................................. 30 3.1 Cell Systems ......................................................................................................................................30 3.2 Cell Culture .......................................................................................................................................31 3.3 Cell proliferation and viability assays ...............................................................................................31 3.4 Drug treatment...................................................................................................................................33 3.5 Na+/K+-ATPase activity screen .........................................................................................................34 3.6 Immunoblotting (IB) .........................................................................................................................34 3.7 Immunofluorescence (IF) ..................................................................................................................35 3.8 Antibodies .........................................................................................................................................35 3.9 Plasmids.............................................................................................................................................36 3.10 Reverse-transcriptase polymerase chain reaction............................................................................36 3.11 Short-interfering RNA (siRNA) ......................................................................................................37 3.13 Transfection of neuroblastoma cells................................................................................................37 3.12 Statistical Analysis ..........................................................................................................................38
Chapter 4 Results ........................................................................................................................ 39 4.1 Digoxin inhibits cell growth of NB TICs and immortalized adherent and non-adherent cell lines..39 4.2 Digoxin induces apoptosis in a dose dependent manner in IMR-5 and NB88R2 .............................40 4.3 Structure-Activity-Relationship Analyses.........................................................................................41 4.4 RIDK34 inhibits cell survival and proliferation of NB TICs ............................................................45 4.5 RIDK34 induces apoptosis in a dose dependent manner in NB88R2 ...............................................45 4.6 Exogenous expression of murine Na+/K+-ATPase alpha one protects against digoxin-induced
inhibition of cell growth in IMR-5 ..........................................................................................................46 4.7 Digoxin inhibits NB TIC growth independently of its effects on [Na+]i and [K+]i...........................47 4.8 Na+/K+-ATPase alpha one subunit knockdown sensitizes IMR-5 and NB88R2 to digoxin growth
inhibitory effects......................................................................................................................................48 4.9 Na+/K+-ATPase alpha one subunit knockdown sensitizes NB88R2 to RIDK34 growth inhibitory
effects ......................................................................................................................................................49 4.10 RIDK34, Digitoxin and Convallatoxin inhibit the Na+/K+-ATPase at concentrations higher than
the inhibitory concentrations for NB cell growth....................................................................................50 4.11 Digoxin increases phosphorylation status of Src and ERK1/2........................................................52 4.12 RIDK34 increases activation of ERK1/2 and at low doses Src kinase ...........................................53
vi
Chapter 5 Discussion .................................................................................................................. 54 5.1 Digoxin and the novel CG analogue RIDK34 inhibit growth of NB cells.......................................54 5.2 Activity of CG-derived sugar groups ...............................................................................................55 5.3 The C19 substituent affects CG potency ..........................................................................................56 5.4 The lactone moiety is essential for the antitumour activity of CGs .................................................57 5.5 Mechanism of CG antitumour activity and the role of the Na+/K+-ATPase ....................................57 5.6 Silencing of ATP1A1 increases susceptibility to digoxin and RIDK34 antitumour activity............59 5.7 RIDK34 is a novel CG analogue with higher potency and displays lower relative inhibition of
Na+/K+-ATPase activity..........................................................................................................................60 5.8 Putative role of the α3 and α4 subunit in CG-induced antitumour activity ....................................63
5.9 Therapeutic rational for CGs in combinatorial therapy....................................................................63
Chapter 6 Conclusion and Future Directions .......................................................................... 65
References .................................................................................................................................... 69
Appendix...................................................................................................................................... 81
vii
List of Tables
Table 1 Children’s Oncology Group (COG) risk stratification schema, by stage 13
Table 2 Tissue distribution of Na+/K+-ATPase alpha isoforms in humans 23
Table 3 Cardiac glycosides and their respective inhibitory and affinity values for
human Na+/K+-ATPase alpha isoforms expressed in P.Pastoris membranes
28
Table 4 Characteristics of cell lines 41
Table 5 Culturing conditions of cell lines 42
Table 6 Conventional CGs and chemotherapeutic agents used for drug efficacy studies 44
Table 7 List of Antibodies used in immunoblotting, and immunofluorescence studies 46
Table 8 Sequence of RT-PCR Primers 48
Table 9 ON-TARGETplus SMART pool sequences 48
Table 10 Structure-Activity-Relationship Analysis 54
Table 11 Comparison of CG serum toxicity and IC50 values for inhibition of NB88R2
growth and Na+/K+-ATPase activity
62
Table 12 Na+/K+-ATPase expression profile in Neuroblastoma Tumour-Initiating Cells 92
Table 13 Reported Mechanisms of CG antitumour activity 94
viii
List of Figures
Figure 1 Structure of the of Na+/K+-ATPase with the obligatory alpha and beta subunits 21
Figure 2 Structure of Cardiac Glycosides 27
Figure 3 Inotropic and Cardiotoxic effects of CGs 31
Figure 4 Na+/K+-ATPase mediated signal transduction 33
Figure 5 Digoxin inhibits neuroblastoma cell growth 51
Figure 6 Digoxin induces apoptosis in a dose dependent manner in IMR-5 and NB88R 52
Figure 7 RIDK34 inhibits neuroblastoma cell growth with moderate selectivity 56
Figure 8 RIDK34 induces apoptosis in neuroblastoma tumour-initiating cells 57
Figure 9 High K+ medium does not protect against digoxin induced inhibition of cell
growth
58
Figure 10 Exogenous expression of cardiac glycoside resistant murine Atp1a1 protects
IMR-5 from digoxin induced inhibition of cell growth
59
Figure 11 Knockdown of the alpha one subunit of the Na+/K+-ATPase increases
susceptibility to digoxin-induced inhibition of cell growth
60
Figure 12 Knockdown of the alpha one subunit of the Na+/K+-ATPase increases
susceptibility to RIDK34-induced inhibition of cell growth
61
Figure 13 Curve fits, and IC50 determination of Na+/K+-ATPase inhibition in HEK-293
cells for Convallatoxin, RIDK34, and digitoxin
62
Figure 14 Digoxin induces Src and ERK1/2 activation 63
Figure 15 RIDK34 induces Src and ERK1/2 activation 64
ix
Figure 16 CGs possibly activate two parallel death pathways in NB TICs 70
Figure 17 Crystal structure of Na+/K+-ATPase with bound ouabain 93
Figure 18 Reaction scheme for RIDK34 93
x
List of Abbreviations
ALK anaplastic lymphoma kinase
Amp ampere
ATP adenosine-5'-triphosphate
ATP1A1 human sodium potassium ATPase alpha 1 polypeptide gene
Atp1a1 murine sodium potassium ATPase alpha 1 polypeptide gene
Bad Bcl-2 antagonist of cell death
BIRC5 baculoviral IAP repeat-containing 5
BDNF brain-derived neurotrophic factor
cAMP cyclic adenosine monophosphate
CG cardiac glycoside
CHD5 chromatin helicase binding domain 5
CNS central nervous system
COG children's oncology group
DAD delayed after depolarization
DM double minute
DOX doxorubicin
EGF epidermal growth factor
EGFR epidermal growth factor receptor
ER estrogen receptor
ERK extracellular regulated kinase
FBS fetal bovine serum
FGF fibroblast growth factor
FISC focus in synthetic chemistry
GBM glioblastoma multiforme
GRB2 growth factor receptor-bound protein 2
HEK-293 human embryonic kidney cell
Il2rg interleukin-2 receptor subunit gamma
Jnk Jun N-terminal kinase
LOH loss of heterozygosity
xi
MKI mitotic-karyorrhectic index
MYCN myelocytomatosis viral related oncogene
Na+/K+-ATPase sodium potassium adenosine triphosphataase/ sodium potassium pump
NAD(P)H nicotinamide adenine (phosphate) dinucleotide
NB neuroblastoma
NB TIC neuroblastoma tumour-initiating cell
NCX sodium, calcium exchanger
NF-kB nuclear factor-kappaB
NGF nerve growth factor
NME1/2 non-metastatic cells 1/2
NSCLC non-small cell lung carcinoma
NTRK neurotrophic tyrosine kinase receptor
PARP Poly-(ADP-ribose) polymerase
PI3-K phosphatidylinositol 3-kinase
PKC protein kinase C
PNS peripheral nervous system
PPM1D protein phosphatase, Mg2+/Mn2+ dependent, 1D
RA retinoic acid
RT room temperature
SCID severe combined immunodeficiency
SDS sodium dodecyl sulfate
SDS-PAGE SDS-polyacrylamide gel electrophoresis
Shc Src homology 2 domain containing
siRNA short-interfering RNA
SKP skin derived precursor
SOS son of sevenless
SR sarcoplasmic reticulum
Src sarcoma kinase
TBS tris-buffered saline
WCE whole cell extract
1
Chapter 1 Review of Literature
1.1 Overview of Neuroblastoma
1.1.1 Introduction Neuroblastoma (NB) is the most common extra-cranial solid tumour of childhood and the age-
dependent rate of incidence is 9.5 cases per million children (1, 2). NB comprises 8% to 10% of
all childhood tumours and accounts for 15% of all paediatric-oncology related deaths (2, 3). It is
a malignancy derived from neural crest cells that would normally give rise to cells of the
sympathetic nervous system. Primary tumours are commonly found in the adrenal medulla and
along various sites in the sympathetic chain. Enigmatically, NB is very aggressive in high-risk
patients with disseminated disease, while in a unique subset of patients under one year with
disseminated disease NB spontaneously regresses (2, 3). Survival rates, particularly for high-risk
patients, have not improved over the past three decades despite intensive combination therapies
(4, 5). Thus, currently applied modalities of treatment are not curative for the majority of high-
risk patients and novel therapeutic strategies are required.
1.1.2 NB treatments, staging and prognosis
Approximately 40% of NB patients present with localized disease (stage 1 to 3 by the
International Neuroblastoma Staging System [INSS]). Prognosis is generally good for localized
disease with survival rates between 70% and 90% (6, 7). Treatment includes tumour excision
and/or adjuvant chemotherapy depending upon the stage and biology of the tumour. However,
more than half of NB patients are diagnosed with the high-risk form of the disease if they present
with widespread metastasis [stage 4] or have localized disease with poor prognostic biological
markers [stage 2 or 3] (Table 1). These patients require comprehensive treatment regimes that
combine high-dose chemotherapy, radiation therapy, stem cell rescue as well as therapy for
residual disease. Nonetheless, current five-year survival rate for high-risk patients with
metastatic disease is below 35% (4, 5).
2
Table 1. Children’s Oncology Group (COG) risk stratification schema, by stage.
Age MYCN status Ploidy Histology Other Risk group
1 Low 2A/2B Not amplified >50% resection Low
Not amplified <50% resection Intermediate
Not amplified Biopsy only Intermediate
Amplified High
3 <547 days Not amplified Intermediate
≥547 days Not amplified Favorable Intermediate
Amplified High
≥547 days Not amplified Unfavorable High
4 <365 days Amplified High
<365 days Not amplified Intermediate
365-547
days
Amplified High
365-547
days
DI=1 High
365-547
days
Unfavorable High
365-547
days
Not amplified DI >1 Favorable Intermediate
≥547 days High
4S <365 days Not amplified DI >1 Favorable Asymptomatic Low
<365 days Not amplified DI=1 Intermediate
<365 days Missing Missing Missing Intermediate
<365 days Not amplified Symptomatic Intermediate
<365 days Not amplified Unfavorable Intermediate
<365 days Amplified High
Abbreviations: DI, diploid. Adapted from Neuroblastoma, Maris 2007 (2).
1.1.3 Genetic Predisposition to NB
Predisposition to NB has been demonstrated with a subset of patients and is consistent with the
Knudson two-mutation hypothesis for the origin of childhood cancers (8). Approximately, 1-2%
of cases are related to known hereditary NB (8). Patients with familial NB are much younger
3
with median age of diagnosis at 9 months compared to median age at diagnosis of 18 months for
sporadic cases of NB (8). Familial NB patients often present with bilateral adrenal or multifocal
primary tumours.
Many neuroblastoma predisposition genes are thought to play an important role in the normal
development of the sympathetic nervous system. PHOX2B is a homeodomain-containing protein
that is involved in the development of the peripheral nervous system. Inactivating mutations of
PHOX2B is commonly associated with sporadic or familial neuroblastoma and with either
congenital central hypoventilation syndrome, Hirschsprung's disease, or in conjunction with both
diseases (9-11). In addition, inactivation of the NF1 gene has been reported in neuroblastoma (12,
13). Thus the genes implicated in the genesis of Hirschsprung disease (RET, EDNRB, EDN3,
GDNF, ECE1, and ZFHX1B), central hypoventilation (RET, GDNF, EDN3, BDNF and
PHOX2B), and/or NF1 may be involved in the initiation or progression of human
neuroblastoma, especially in cases of coincidence with these syndromes (12, 13).
Recently, activating mutations in the tyrosine kinase domain of the anaplastic lymphoma kinase
(ALK) oncogene have been reported in neuroblastoma. These germ-line mutations are thought to
account for most cases of hereditary neuroblastoma and encode a single-base substitution in
critical regions of the kinase domain resulting in constitutive activation of the kinase (12). Single
base pair mutations or amplifications that result in oncogenic activation of ALK are somatically
acquired in 5% to 15% of neuroblastoma (12, 14, 15). The identification of these and other
familial neuroblastoma predisposition genes will benefit affected families, as well as sporadic
neuroblastoma patients by contributing to our understanding of the initiation and progression of
NB. In addition, although ALK and PHOX2B mutations account for the majority of familial
cases of neuroblastoma, additional familial genes may still be discovered.
1.1.4 Genetic Characteristics of NB
Near diploid and triploid DNA content
The DNA content of neuroblastoma cells can be classified into near-diploid and near-triploid
(hyperdiploid). Near-diploid tumours are usually associated with unfavourable prognosis while
hyperdiploidy is more likely to be detected in favourable tumours (16, 17). DNA ploidy is a
useful prognostic marker for patients less than one year of age with advanced disease and is
4
currently used by the Children’s Oncology Group (COG) and other international groups for
determining risk-adapted therapy.
Amplification of MYCN Locus at 2p24
MYCN amplification is the best-characterized genetic abnormality in neuroblastoma.
Amplification occurs in approximately 20% of primary tumours and is strongly correlated with
advanced stage disease and poor patient outcome (18-20). MYCN belongs to the MYC family of
transcriptional factors and maps to chromosome position 2p24. In NB, MYCN is commonly
amplified and found on extrachormosomal double minutes (DM) or in homogenously staining
regions in the genome distinct from 2p24 (21, 22). In addition besides MYCN, large regions
flanking either side of this gene including ALK are amplified. A recent study has suggested that
amplification of a region proximal to MCYN and independent of both ALK and MYCN may
also have prognostic significance (23).
1p and 11q loss of heterozygosity
Deletion at 1p36 and of 11q occurs in 25-30% and 35-45%, in primary neuroblastoma tumours
respectively at diagnosis (24). Deletion at 1p36 is commonly observed in patients with advanced
disease and is often associated with MYCN amplification and other poor prognostic features (2,
25, 26). Also it appears that 1p36 deletion is associated with increased risk of disease relapse in
patients with localized tumors. Unlike 1p deletions, 11q loss of heterozygosity (LOH) is very
rarely observed with MYCN amplification yet is still associated with high-risk features and is an
independent indicator for tumour relapse. The prognostic value of 1p and 11q for predicting
disease relapse has led to the incorporation of these markers in patient risk stratification and the
determination of appropriate therapy. Although there are tumour suppressors located in these
regions, including p73 and Chromatin Helicase Binding Domain 5 (CHD5), research is ongoing
to identify novel tumour suppressors at 1p36 and 11q involved in the pathogenesis of
neuroblastoma (27).
17q addition and other cytogenetic sites
Gain of 17q genetic material is a predictor of poor outcome and occurs in over half of the
patients with NB, making it the most common cytogenetic abnormality in NB (28, 29). MYCN
and/or 1p deletion occurs in 75% of tumours with 17q addition. Unbalanced translocation of 1p
and 17q resulting in addition of distal 17q and loss of distal 1p genetic material frequently occurs
5
in NB. 17q translocations are also thought to occur at other genetic regions in addition to 1p.
Several candidate genes conferring a selective advantage at the distal 17q site for tumour cells
include NME1/2 PPM1D, and BIRC5 (survivin); however, the role of these and other candidate
genes in the initiation and progression of NB is still poorly understood (28).
1.1.5 NB Characteristics
Neurotrophin Receptors: TrkA, TrB and TrkC
The neurotrophin receptors (NTRKA, NTRKB and NTRKC encoding TrkaA, TrkB and TrkC) and
their ligands (NGF, BDNF and neurotrophin-3, respectively) can induce survival and
differentiation of neural cells including sympathetic neuroblasts (30). High expression levels of
TrkA and its ligand NGF correlates with younger onset of NB, lower risk stage, the absence of
MYCN amplification and is associated with favourable outcome (31). In addition, NB cells
expressing high levels of TrkA undergo differentiation in the presence of NGF or programmed
cell death when deprived of NGF; conversely, high levels of TrkB and its ligand BDNF is
associated with MYCN amplification and poor outcome (2, 32). Evidence suggests that co-
expression of TrkB and BDNF activate an autocrine or paracrine survival pathway that promotes
chemotherapy resistance and metastases, possibly by suppressing anoikis (loss of cell/cell and
cell/extra cellular matrix attachment) resulting in cell death (33, 34). Investigation into targeted
therapies are ongoing and are promising such as with the Trk selective inhibitor Lestaurtinib,
which is in preclinical and phase I clinical study (35).
Tumour Histology
The degree of neuroblastic maturation toward a mature ganglion phenotype correlates with
favourable diagnosis. Tumours with largely undifferentiated morphology have poor prognosis,
while tumours containing Schwann, chromaffin and neuronal cell types have more favourable
prognosis (36). These findings allowed Shimada and colleagues to develop a schema that was
incorporated into the International Neuroblastoma Pathology Classification System (37).
1.1.6 Neuroblastoma Tumour-Initiating Cells
Support for Tumour-Initiating Cells in Neuroblastoma The clinical presentation and treatment response of neuroblastoma supports the likelihood of a
6
cell population with stem cell-like as well as tumour-initiating properties. In NB, histology is one
method used to assess patient risk (see Section 1.1.5) and would suggest that NB is derived from
a neural crest or sympathoadrenal precursor. NB tumors are comprised of both undifferentiated
tumour cells that express neural stem cell markers such as nestin, and more differentiated cells
with markers of mature neural cells and Schwann cells. This heterogeneity is consistent with the
existence of a cancer stem cell model with hierarchical properties. Moreover, in vitro neuronal
induction of immortalized neuroblastoma cells with retinoic acid (RA) and/or dibutyryl cyclic
AMP gives rise to cells with neuronal (38), chromaffin (39) or Schwannian (40) phenotypes,
further supporting the existence of a cell with cancer stem cell properties in NB.
Cancer Stem and Tumour-Initiating Cell Theory One possible reason for the ineffectiveness of current treatments for high-risk NB is the lack of
cell culture and in vivo models for metastatic NB that can be utilized for drug testing. Recently,
NB cells with cancer stem cell and tumour-initiating properties have been identified that may be
used towards better understanding and developing novel therapies against NB. Evidence suggests
that cancer stem cells exist as a subpopulation and share essential features with normal stem cells
such as the expression of stem cell genes and self-renewal capacity in cell culture (41-43). In
addition, these cells exhibit enhanced tumour-initiating capacity as compared to most
established-adherent cell lines grown in serum and are often organized as a hierarchy, with
cancer stem cells able to generate both a differentiated daughter cell that is non-tumourigenic,
and a tumourigenic copy of itself (41-43). Thus far, cancer stem cells have been described in
adult leukemia (44), breast cancer, pediatric brain tumors (45, 46), melanoma (47),
ependymomas (48), colon cancer (49) , and head and neck squamous cell carcinoma (50).
Characterization of NB Tumour-Initiating Cells
NB tumour-initiating subpopulations may be isolated from established cell lines or cells taken
directly from primary tumour or bone marrow metastasis samples (51-53). Sub-populations, have
been idenifited by various groups through the use of sorting markers and functional assays (51-
53). Mahller et al. were able identify tumour-initiating cells from the established adherent cell
line LA-N-5, which in serum free conditions grows as spheres and express CD133+ (52). In
contrast Das et al. were able to increase tumour stemness in the adherent established cell line
SK-N-(BE-2) by identifying side-population (SP) cells, which are cells that negatively stain for
Hoechst dye (53). This side-population exhibited even greater tumour-initiating capacity upon
7
further selecting SP cells that migrate towards the media of stressed bone marrow stromal cells
(53). These cells were identified as SPm cells, which displayed further tumour stemness by
subjecting these cells to oxidative stress and hypoxia followed by reoxygenation and were
identified as SPm(hox) cells (53). These functional assays increased the tumour-initiating capacity
of SK-N-(BE-2) derived cells, as only 100 SPm(hox) cells were required to form tumours in
BALBc, nude/nude mice, unlike SPm and normal SK-N-(BE-2) that require 1 ×104 and 2.5 ×105
cells respectively to be injected to form tumours (53).
In contrast to the previously mentioned cell systems, the Kaplan laboratory used primary cells
that were not grown at any point in serum conditions in order to identify neuroblastoma tumour-
initiating cells (NB TICs). Given that the most common site of metastasis is the bone marrow,
the majority of NB TICs were isolated from bone marrow aspirates of high-risk patients, using a
filtration technique to enrich for NB cells that tend to form aggregates in this niche (51). The
dissociated cells reformed spheres in neural crest stem cell media, expressed neuroblastoma
markers such as NB84 and tyrosine hydroxylase, neural crest precursor markers such as nestin,
and had chromosomal alterations consistent with NB (51). Similar to cancer stem cells, sphere-
forming neuroblastoma cells were capable of self-renewal and differentiation into the cell types
observed in NB (51). Also, sphere forming neuroblastoma cells, when injected orthotopically
into the adrenal fat pad of immunocompromised mice, formed metastatic tumours that could
serially reform spheres and could recapitulate the original phenotype of the tumour (51). In
addition, the frequency of tumourigenic cells is much higher in NB TICs as only 10 NB TICs
injected into SCID/Beige mice are required to form tumours which is several orders of
magnitude less than immortalized NB cell lines (51-54).
Also unlike cancer stem cells from leukemia, colon and brain tumors, spheres from NB bone
marrow metastasis do not appear to contain a hierarchical organization, in which a subpopulation
of tumour cells express a distinct cell-surface marker and is capable of giving rise to all NB
subtypes (51). This could be because a prospective marker has yet to be identified for the TIC
population in NB, or that there is a fundamental difference between developmental tumours such
as NB relative to adult tumours, or possibly because metastatic neural crest-derived cells are
relatively homogenous regarding TIC potential (51). Recently, Morrison and colleagues
demonstrated that single unsorted melanoma cells are able to form a tumour in SCID/Beige
(Il2rg-/-) mice. They suggested that at least in melanoma, tumourigenic cells are very common
with little evidence implicating hierarchical organization in these tumours (55). Similarly, it is
8
possible that neuroblastoma cells that metastasize to the bone marrow are highly enriched in
tumour-initiating potential compared to the tumour from which they arose and do not exhibit
hierarchical organization in this niche (51).
Applications of NB TICs There are several potential applications for NB TICs. In one case, neuroblastoma tumour sphere
cells isolated from a patient in remission and showing no morphologic evidence for disease was
able to form tumours in immunocompromised mice. As anticipated from these results this patient
clinically relapsed a few short months after the tumour spheres were isolated. Given that NB
TICs can be isolated from the bone marrow of patients in relapse as well as in remission with a
high chance of relapse, suggests that NB TICs may in the future be used as a predictor of clinical
behaviour, treatment efficacy in high-risk patients, and in detecting minimal residual disease. By
comparing the biology of NB TICs at different stages of treatment from newly diagnosed
patients, patients in remission and to relapse patients, we may better understand the cellular and
molecular events leading to relapse and disease progression in neuroblastoma. Lastly, current
therapies in NB patients are largely ineffective, and while high-risk patients may go into
remission, they invariably relapse. Given their potent tumour-initiating capacity, NB TICs likely
play an integral role in treatment response and failure. Thus drugs that specifically target NB
TICs are more likely to be effective agents for treating NB. As described in the proceeding
sections, cardiac glycosides, which are inhibitors of the sodium potassium pump represent a
novel class of agents with efficacy against NB TICs.
1.2 Sodium Potassium ATPase
1.2.1 Introduction
The sodium potassium pump (Na+/K+-ATPase or sodium pump) belongs to the P-type ATPase
family and is responsible for the low intracellular ratio of Na+ to K+ ions. The Na+/K+-ATPase
from the hydrolysis of ATP, transports three Na+ ions out of the cell in exchange for two K+ ions
into the cell (56-58). In mammals, the Na+/K+-ATPase regulates several vital cellular functions;
specifically, it regulates ion homeostasis, cell volume, acts as a signal transducer and maintains
resting membrane potential, which is essential to the function of electrogenic cells such as
cardiac myocytes and neurons (58-62). The sodium pump is also essential for the function of
9
secondary transporters that are coupled to the Na+ gradient across the plasma membrane such as
the Na+/Ca2+ exchanger (56-58). Thus, as a result of ion regulation and ATP hydrolysis, the
Na+/K+-ATPase regulates much of the cellular metabolic rate and accounts for 1/3 to 2/3 of
energy expenditure depending on the cell type (63-65).
1.2.2 Structure and Function of the Na+/K+-ATPase
The Na+/K+-ATPase consists of an α and β subunit in equimolar ratio (Fig. 1). The α subunit is
the catalytic polypeptide and has ten transmembrane domains (63). The α subunit contains the
extracellular binding sites for K+ ions and cardiac glycosides (CGs). On the intracellular face of
the α subunit are the Na+ binding and phosphorylation sites (65). The activity and expression of
the catalytic α subunit is regulated by the β subunit, which is a single transmembrane polypeptide
and contains a highly glycosylated extracellular loop (65). The β subunit is also thought to act as
a chaperone protein for localization of the α subunit to the plasma membrane (63-65). The third
subunit is the FXYD subunit, which is not required for the function and localization of the
Na+/K+-ATPase; however, the seven FXYD isoforms appear to alter the Na+/K+-ATPase affinity
for K+, Na+ and ATP (66-68).
The Na+/K+-ATPase like other ion-transporting ATPases has two major conformations, denoted
as E1- and E2- (69). According to the Albers-Post model, in the presence of K+, Na+, Mg2+ and
ATP, the α subunit becomes phosphorylated, which is followed by occlusion of three Na+ ions.
This high energy E1P state of the Na+/K+-ATPase undergoes a conformational change to the
lower E2P form, which translocates and releases Na+ ions into the extracellular milieu (69-71).
The presence of K+ ions causes the E2P form to become dephosphorylated followed by binding
of K+ ions. K+ binding leads to the E2- to E1- transition, which is hastened by the presence of
ATP (69-71). Upon transition to the E1- conformation, K+ is released from the Na+/K+-ATPase
into the cell and ATP binds to repeat the cycle (69-71).
10
Figure 1. Cartoon structure of the Na+/K+-ATPase with the alpha subunit, beta subunit and FXYD domain. Binding sites for ATP and cations are in the second intracellular loop (i) and ouabain is bound deep inside the membrane approximately at the M4-M6 helices (ii) (66).
1.2.3 Subunit Diversity and Expression of the Na/K-ATPase
There are four α subunits, α1-α4, and three β subunits, β1-β3 (63-65, 72, 73). The primary
sequences of the four α and three β subunits are highly conserved among species (63). The
individual α and β isoforms in rat, mouse and human show greater than 90% sequence identity;
the only exception is β3 with 75% similarity across species (as reviewed in 65, 66, 73). Despite
this, very modest changes to the primary structure of the Na+/K+-ATPase can result in
extraordinary changes in the pharmacological properties of the α and β isoforms across species.
For example, sheep, rat and human sequences differ only slightly in the α1 subunit of the
Na+/K+-ATPase; however, just two single residue substitutions (glutamine-118 to arginine and
aspargine-129 to aspartic acid) in the α subunit increases ouabain resistance 1,000 fold in rodents
relative to sheep and human (74, 75).
11
There is slightly less similarity among the four α or the three β subunits within a given species.
In mammals, the α1, α2 and α3 share ~87% sequence homology, while the α4 is approximately
76%-78% similar to the other isoforms (65, 66, 73). The β isoforms are more divergent than the
α subunits with β2 and β3 sharing with β1 34% and 39% similarity respectively, whereas β2 and
β3 have 49% amino acid identity (64).
The Na+/K+-ATPase is most abundantly expressed in ion-transporting epithelia such as the
kidney and in excitable tissues like the brain, skeletal muscle and cardiac muscle. Alternatively,
the Na+/K+-ATPase is weakly expressed in vascular smooth muscle whereby electrochemical
gradients are less important for physiological function. In addition to differential expression, the
α and β isoforms of the Na+/K+-ATPase are expressed in a tissue-specific manner (Table 2). The
α1 and β1 isoforms are expressed ubiquitously and the α1/β1 complex is the major isozyme in
nearly every tissue (Table 2). Expression of α2 is mainly restricted to the adult heart, vascular
smooth muscle, cartilage, bone adipocytes and various regions of the brain. Thus far, α3 is
mostly localized to the conductive system of the heart and the central and peripheral nervous
system (76-80). Unlike α1, α2 and α3, expression of α4 is mainly restricted to the testis (81).
The β2 isoform expression is restricted to the heart, brain, cartilage and erythrocytes, whereas β3
isoform expression is restricted to lung, brain, cartilage and erythrocytes. Not surprisingly, the
tissue restricted expression of the α and β isoforms and their respective affinities for Na+ ions, K+
ions and CGs in large part determines the physiologic response of tissues to these substrates (81-
83).
12
Table 2. Tissue distribution of Na+/K+-ATPase alpha isoforms in humans.
Na+/K+-ATPase isoform Gene Tissue Distribution
Alpha 1 ATP1A1 Ubiquitous
Alpha 2 ATP1A2 Skeletal muscle, heart, brain, adipocytes, cartilage, bone, vascular smooth muscle, eye
Alpha 3 ATP1A3 Neurons (PNS and CNS), ovaries, white blood cells, heart (conduction system)
Alpha 4 ATP1A4 Testis
Beta 1 ATP1B1 Ubiquitous
Beta 2 ATP1B3 Heart, brain, cartilage, erythrocytes
Beta 3 ATP1B3 Lung, brain, cartilage, erythrocytes
1.2.4 Regulation of the Sodium Potassium ATPase
Several hormones including endogenous cardiac glycosides (CGs), catecholamines and various
other steroid or peptide hormones regulate the Na+/K+-ATPase through short-term or long-term
regulatory mechanisms. Short-term regulation involves; 1) direct affects on the kinetic rates of
the enzyme or 2) translocation of formed oligomers of the pump between intracellular
compartments and the plasma membrane (83-86). Modulating its affinity for substrates in part
regulates kinetic rates of the Na+/K+-ATPase. For example, Ser-16 phosphorylation of the α
subunit by PKC can increase the pump’s affinity for Na+ ions (87). In addition, in alveolar lung
epithelium, dopamine regulates Na+/K+-ATPase activity by promoting recruitment of the pump
to the basolateral membrane from intracellular stores (88).
Long-term regulation of the Na+/K+-ATPase occurs by transcriptional and post-transcriptional
mechanisms including changes in transcription rate, mRNA stability, translation rate, protein
degradation and isozyme expression (reviewed in 86, 89). Gene regulation occurs in an isoform
specific manner as individual isoforms of the α and β subunits are encoded by different genetic
loci. During late neonatal stages cardiac tissue switches expression from α1 and α3 mRNA to
high levels of mostly α1 and α2 transcripts due to isoform-specific regulation by thyroid
13
hormone and glucocorticoids (90). Also aldosterone regulates isoform expression, as it can
induce α3 expression in the brain, while targeting α1 expression in the kidney (86, 90). Several
hormones, including thyroid hormone, aldosterone and glucocorticoids, regulate the spatial-
temporal expression of the Na+/K+-ATPase leading to its very complex isozyme expression,
which likely explains the different physiologic response of tissues to Na+/K+-ATPase substrates,
namely Na+ ions, K+ ions and CGs.
1.2.5 The Na+/K+-ATPase: A Novel Anticancer Target
Previous reports have correlated increases in Na+/K+-ATPase activity with tumour progression
and in some cases increased Na+/K+-ATPase activity has been observed prior to morphological
evidence of an established tumour (91-93). Correlation between elevated Na+/K+-ATPase and
tumour growth would suggest that the Na+/K+-ATPase likely plays an important, yet poorly
defined role in initiation and progression of tumours (91-93). Increased Na+/K+-ATPase activity
has been theorized by Weidemann to be a compensatory mechanism to higher Na+ content that
results from increased passive Na+ membrane permeability of neopastic cells (94). Collectively,
the higher intracellular Na+ content and higher Na+/K+-ATPase activity in tumour cells facilitate;
1) an increase in co-transport and accumulation of nutrients for the metabolic demands of
transformed cells; and 2) support cell proliferation as a result of higher intracellular K+
concentration and lower cAMP from increased pump activity (94).
Increased activity of the Na+/K+-ATPase may be accomplished by increased isozyme expression
of the Na+/K+-ATPase at the plasma cell membrane. The α3 Na+/K+-ATPase has higher affinity
for K+ ions and lower affinity for Na+ ions relative to other isoforms, and functions with lower
occupancy of the intracellular Na+ sites (95). This may suggest that elevated expression of α3 in
tumours can manifest high Na+ and K+ in transformed cells as predicted by Weidman. Indeed,
reports have demonstrated an increase in α3 expression in kidney, pancreatic and bladder cancer
with associated decreases in α1 expression in both colorectal and prostate cancers (96, 97).
Alternatively, elevated α1 expression has been observed in both non-small cell lung cancer
(NSCLC) and glioblastoma multiforme (GBM) (98, 99). In these tumours the expression of α2
and α3 subunits at the plasma membrane did not reveal any significant differences compared to
respective controls (normal lung tissue and normal brain tissue). Elevated expression of α1
and/or α3 subunits in some tumours may account for overall increases in Na+/K+-ATPase
14
activity. How increased expression of α subunits may drive the development of tumours is still
poorly understood and continued investigation is required to elucidate the effects of the α1 and
α3 subunits on tumour growth.
Specific cancer related changes are also observed in the β subunit and FXYD domain of the
Na+/K+-ATPase. Rajasekaran et al. demonstrated that decreased expression of the β1 subunit
correlated with poorly differentiated cancer cell lines derived from colon, prostate, kidney and
pancreatic cancer (100-103). Furthermore, they suggested that the Snail transcription factor
inhibits gene expression of the β1 subunit and cadherin, leading potentially to an epithelial-
mesenchymal transition permitting cancer cells to become more invasive (103).
Several FXYD polypeptides are specifically associated with cancer and almost exclusively
expressed in cancer cells. For example, the FXYD3 (Mat-8: a mammary tumour marker) is
highly up-regulated in prostate, colorectal, pancreatic, gastric (adenocarcinoma) and breast
tumours (104-108). Yet, FXYD3 overall appears to be down-regulated and sometimes mutated in
various types of lung cancers including, adenocarcinoma, squamous cell carcinoma, large cell
carcinoma and small cell carcinoma (109). This down-regulation in lung cancer cells may be the
resultant loss of an epithelial phenotype, as over-expression of wild type FXYD3 in tumour cells
restores the well-demarcated distribution of cortical actin that separates the apical layer from the
basolateral regions of epithelial lung cells (109). Furthermore, a recent study on intestinal
suggested that the interaction of FXYD3 with the Na+/K+-ATPase may indirectly promote tight
junctions and cortical actin leading to the polarization and differentiation of intestinal epithelia.
Still the specific role of FXYD3 in cancer is poorly understood yet it appears that its loss may
promote an undifferentiated state in tumour cells.
Several pieces of evidence show that targeting the Na+/K+-ATPase may have anticancer effects
including; (1) its role in cell adhesion and polarity, (2) observations of increased expression in
several cancer models and (3) observations of its role in signal transduction (see section 1.3.5).
Given the potential role of the Na+/K+-ATPase in cancer, specific targets of the pump such as
CGs may be predicted to have therapeutic value.
15
1.3 Cardiac Glycosides
1.3.1 Introduction
The membrane inserted Na+/K+-ATPase is the known target of cardiac glycosides (CGs). CGs or
cardiotonic steroids have long been used for a variety of conditions. Its best described therapeutic
application is for the treatment of cardiac dysfunction namely, congestive heart failure and atrial
arrhythmias (110, 111). These inotropic agents are thought to increase intracellular calcium as an
indirect effect of inhibiting the Na+/K+-ATPase. Retrospective, epidemiological studies of breast
cancer patients would suggest that therapeutic concentrations of CGs could be achieved to treat
breast carcinoma with minimal toxicity (112, 113). This and other findings thereafter has lead to
the emergence of CGs as a possibly novel drug target for the treatment of cancers.
1.3.2 Chemical structure
CGs are compounds with three basic structures: 1) a tetracyclic steroid ring (steroid nucleus); 2)
a sugar moiety at position C3 of the steroid nucleus and 3) a lactone moiety at the C17 position
in the steroid nucleus (Fig. 2). The lactone moiety defines two classes of CGs. The lactone group
of cardenolides is an unsaturated butyrolactone ring and for bufadienolides this is a α-pyrone
ring. CGs derived from Digitalis and Nerium plant species (digoxin, digitoxin) have A/B rings in
cis conformation while the B/D rings are in trans conformation resulting in the ‘U’ shape of these
compounds. A number of CGs produced by the milkweed family Asclepiadacea, including
calactin, uscharin, 2″-oxovoruscharin and the derivative UNBS1450 contain A/B fused rings in
trans confirmation giving rise to a flat shape of the nucleus structure as well as a unique
dioxanoid attachment of the sugar moiety to the steroid nucleus (double link) (83, 114, 115).
16
Figure 2. Structure of CGs. CGs are composed of a tetracyclic steroid ring, a lactone ring and sugar moiety. CGs are classified as bufadienolides or cardenolides based on the structure of the lactone ring. Digitoxigenin is the aglycone of digitoxin as it lacks three digitoxose sugar groups and digitoxin lacks the hydroxyl group attached to the C12 carbon of digoxin. Cardenolides are shown with the exception of bufalin (bufadienolide) and oleandrin (enolide).
1.3.3 CG binding site on the Na+/K+-ATPase
Recently, the crystal structure of the Na+/K+-ATPase alone as well as with bound K+ and ouabain
has been resolved from shark rectal tissue (see Appendix Fig. 16). The lactone ring and the
tetracyclic steroid ring are thought facilitate binding to the M4-M6 helices of the Na+/K+-ATPase
(116). The carbonyl group in the lactone ring is predicted to disrupt the K+ coordination site II of
the Na+/K+-ATPase and is thought to cause a structural change in the Na+/K+-ATPase, leading to
unwinding of the M4 helix creating a hydrophobic binding cavity for the steroid nucleus (116).
The steroid nucleus and lactone group likely cooperate to stabilize binding interactions with the
Na+/K+-ATPase; as extensive experiments with simple lactones alone lack inotropic activity of
digoxin, which suggests that the lactone group alone is not sufficient for binding to the pump
(117, 118).
The proximity and distribution of the CG binding site around the K+ coordination site, may in
17
part explain the ability of K+ ions to antagonize the inhibitory effects of CGs on the Na+/K+-
ATPase (116). One implication of this property is that CGs likely bind with high affinity to the
Na+/K+-ATPase in unbound potassium states (e.g. E2P) and bind with low affinity to potassium
bound states of the Na+/K+-ATPase (E2-K+Pi) (74, 116, 119).
Furthermore, Paula et al. suggested that the sugar moiety may play an important role in CG
binding, with minimal effects on Na+/K+-ATPase inhibition. This is readily demonstrated with
ouabain and ouabagenin which are similar in structure except for a hydroxyl group in ouabagenin
where the rhamanose sugar binds to the tetracyclic steroid ring of ouabain (119). These two CGs
have nearly identical inhibitory potencies on the Na+/K+-ATPase; however, the rhamnose sugar
of ouabain is predicated to form hydrogen bonds with the M4 helix which confers ~300 times
great binding affinity than ouabagenin (116, 119). In addition, digitoxin and its aglycone
digitoxigenin (substitution of digitoxose sugar moiety for a hydroxyl group), displays a similar
disparity between binding affinity and the Na+/K+-ATPase inhibitory values (Table 3).
Table 3. CGs and their respective affinity and inhibitory values for human Na+/K+-ATPase alpha isoforms expressed in P.Pastoris membranes (from Katz et al.120).
Cardiac Glycoside Binding affinity Kd (nM) Inhibitory values Ki (nM)
α1 α2 α3 α1 α2 α3
Digoxin 87±6 25.6 ±2.8 25 ±2.4 250 ±44 63 ±10 136 ±24
Digitoxin 38 ±3 18.3 ±4 14±3.9 72 ±8 23 ±4 52 ±11
Digitoxigenin 270±21 322±11 310.4±27 76±18 45 ±6.7 72.7 ±7
Ouabain 9.8 ±. 33 21.9 ± 0.56 11.1 ±1.3 97±4.6 94 ±16 97 ±13
1.3.4 CGs: Inotropy and Adverse Effects
CGs, are characterized as inotropic agents because they increase the rate and force of cardiac
contraction. These properties of CGs support the use of digoxin and digitoxin in the treatment of
cardiac dysfunction. CGs are often used only when standard therapy including angiotensin-
converting enzymes and β-blockers fail to relieve symptoms of heart failure (121). One reason
for caution is that their upper therapeutic concentrations overlap with symptoms of CG-induced
18
toxicity of the central nervous system (CNS) namely depression, decreased vision acuity and
psychosis (122). Also given the high levels of Na+/K+-ATPase expression in epithelia tissue, it is
not surprising that there are detrimental gastrointestinal effects from CG therapy. Adverse
effects, such as diaherra, nausea and vomiting, are thought to arise from increased motility of the
GI tract (123,124).
The most widely supported theory to explain the toxic and inotropic effects of CGs is the Na+-lag
hypothesis (Fig. 3). Under normal physiologic conditions, the Na+/Ca2+ exchanger bound to the
plasma membrane utilizes the Na+ gradient established by the Na+/K+-ATPase to drive the efflux
of Ca2+ ions. The inhibitory effects of CGs on the Na+/K+-ATPase results in a slow increase of
intracellular Na+ concentrations, resulting in a dampening of the Na+ gradient and an increase of
intracellular Ca2+ due to diminished Na+/Ca2+ exchanger activity. The excess intracellular Ca2+ is
then taken up into the sarcoplasmic reticulum (SR) by the SR Ca2+-ATPase, increasing
intracellular stores, which when released, strengthens subsequent contractions of affected cardiac
myocytes. In addition to mechanical effects, CGs affect the electrical properties of cardiac tissue
through; 1) indirect autonomic action that decreases conduction and nerve impulses to the heart,
and 2) direct actions on cardiac cells leading to shortening of the plateau phase and an overall
shorter action potential (123). Partial pump inhibition in cardiomyocytes (roughly 30%) is
believed to mediate the therapeutic effects of CGs, while greater than 50% inhibition leads to
toxicity (122). At higher concentrations further decreases in the Na+ gradients reduces resting
membrane potential. With a more neutral (less polar) membrane potential, delayed after
depolarizations (DADs) are more likely to occur following a normal action potential. DADs are
associated with overloading of the intracellular calcium stores in the SR and can lead to their
own action potential and cardiac myocyte contraction. If DADs are coupled with normal action
potentials in the Purkinje conduction system of the heart, bigeminy (paired heart contractions)
will result. DAD evoked action potentials can themselves induce a DAD and ectopic beats
causing tachycardia that may ultimately deteriorate into ventricular fibrillation, which can be
fatal (123).
Although this theory is well accepted, there are questions with regards to the exact mechanism of
Ca2+ regulation by the Na+/K+-ATPase and the specific roles of the α1, α2 and α3 isoforms
expressed in the heart. Studies from α1 -/+ and α2 -/+ heterozygous mice as well as evidence of
subcellular colocalization of α2 with the Na+/Ca2+ exchanger in the plasma membrane nearby the
19
SR suggests that α2 sensitivity could mediate inotropic effects of CGs (124). Furthermore, in this
model the α1 isoform, based on its uniform distribution in cardiac myocytes would be
responsible for bulk transport of Na+ and K+ ions into and out of the cell. Inhibition of the α1
Na+/K+-ATPase isoform would therefore affect global cellular ion concentration giving rise to a
diminished electrochemical gradients with possible toxic outcomes (124). The historical
assumption is that partial inhibition versus complete inhibition of the Na+/K+-ATPase gives rise
to the respective therapeutic and toxic effects of CGs; however, the differential expression of
isoforms in the heart may also in fact explain these differences. In particular the α2 subunit may
mediate inotropic effects of CGs by close association with the SR allowing for subcellular Ca2+
oscillations, whereas inhibition of all isoforms including α1 may result in whole cell increases of
Na+ ions leading to cardiotoxicity (124). Previously, drugs designed to increase inotropic activity
while minimizing cardiotoxicity attempted to increase the binding affinity of CGs and decrease
inhibition of these drugs, based on the assumption that not all of the inotropic effects of CGs are
the result of Na+/K+-ATPase inhibition (124). However, if true this relatively new theory for the
therapeutic mechanism of CGs would shift the focus to design novel agents that preferentially
target isoforms other than α1 that may improve inotropic effects and decrease cardiotoxicity.
This same rationale may also be employed in drug development of CGs with anticancer potential
(see Section 5.8).
20
Figure 3. Inotropic and cardiotoxic effects of CGs. According to the Na+-lag hypothesis, partial inhibition of the Na+/K+-ATPase by CGs leads to a transient increase of [Na+]i. This in turn leads to an increase in [Ca2+]i via a reduction in Na+/Ca2+ exchanger (NCX) activity. Excess [Ca2+]i is taken up by the endoplasmic reticulum and upon stimulation, release of stored Ca2+ amplifies the normal force of contraction. Complete inhibition of the Na+/K+-ATPase by CGs evokes a constantly elevation in [Ca2+]i that overwhelms the sarcoplasmic reticulum and leads to deleterious cardiotoxic effects, including hypertrophy and tachycardia.
1.3.5 The role of Na+/K+-ATPase in CG Antitumour Activity
In comparing a series of compounds derived from the CG 2”-oxovoruscharin, Van Quaquebeke
et al. found that compounds with very poor inhibition of the Na+/K+-ATPase (IC50 >50 µM) did
not inhibit cancer cell growth (IC50 >10 µM) (114). This association might suggest that inhibition
of cancer cell growth by CGs requires inhibition of the Na+/K+-ATPase. These conclusions were
drawn by Winnicka et al. following 24-48 hour treatment with ouabain (100 nM and 200 nM),
that induced apoptosis and a sustained increase in intracellular Ca2+ in human breast cancer cells
(MDA-MB-435, see Appendix Table 13) (125). Furthermore, the doses used to induce apoptosis
(1 µM and 10 µM) in prostate cancer cells (PC-3, LNCaP and DU145) also were associated with
sustained elevation in intracellular Ca2+ (126). These studies in breast cancer and prostate cancer
models support the following interpretation: inactivation of the Na+/K+-ATPase and inhibition of
Na+/K+-ATPase ion regulation induces antitumour activity.
21
Although, Na+/K+-ATPase and cancer growth inhibition are associated, this alone does not
completely explain the antitumour activity observed with CGs. In many cases the concentration
of cardiac glycoside used to induce antitumour activity do not inhibit the Na+/K+-ATPase and
pump-mediated intracellular concentrations of Na+, K+ and Ca2+. For example, ouabain (10-100
nM) induces antiproliferative effects with less than 25% of pump inhibition (over a period of
seven days) (127). In this same study, associated increases in levels of p21, Src, ERK and JNK
signaling were reported (61). In SH-SY5Y neuroblastoma cells, apoptosis was observed with
ouabain (<100nM) and was associated with activation of ERK, p53 and pBad at lower
concentrations than is required to dramatically increase intracellular Ca2+ (128). Thus, evidence
indicate that both an ionic pathway and a non-ionic signaling pathway play important roles and
work possibly in parallel to induce anticancer activity in addition to likely other Na+/K+-ATPase-
dependent functions such as inotropy, and internalization of the Na+/K+-ATPase itself.
Na+/K+-ATPase -mediated Signal Transduction
Recently recognized is the capacity of CGs to induce, downstream of binding to the Na+/K+-
ATPase, a myriad of signaling proteins that leads to changes in gene expression, and the rate of
proliferation in a manner that varies betweens cells, the concentration of CG and even with
individual CGs (Fig. 4) (129-137). As described in the previous section, the signaling pathways
activated in response to CGs do not necessarily induce large changes in cytosolic Na+ ions (61,
128, 138). In part this may be explained by two pools of the Na+/K+-ATPase within the plasma
membrane; one pool being the classical energy transducing ion pump and the other, the signal
transducing form of the pump restricted to mostly caveolae (134). Although the precise
mechanism by which the Na+/K+-ATPase induces a signaling cascade has not been completely
elucidated, the following sections describe the current knowledge of CG-induced signaling via
the Na+/K+-ATPase.
22
Figure 4. Na+/K+-ATPase-mediated signal transduction. Upon binding to the Na+/K+-ATPase, CGs induce a signaling cascade. CGs release Src from the inhibitory restraint of the Na+/K+-ATPase, allowing Src to activate EGFR and the Ras-MEK-ERK1/2 pathway. Also CG binding activates a complex consisting of PLC tethered to the Na+/K+-ATPase and the IP3 receptor on the endoplasmic reticulum. By binding to the Na+/K+-ATPase, CGs increase IP3 formation, which in turn increases intracellular stores of Ca2+. These two signaling pathways can affect PI3K-Akt signaling, NF-κB signaling and a host of other signaling cascades leading to a proliferative or antiproliferative response depending on the cell type, CG concentration, and CG (129-137).
Src-EGFR-Ras-ERK1/2 signaling The Na+/K+-ATPase undergoes a conformational change, following ouabain binding, that
activates the cytoplasmic tyrosine kinase SRC in a complex that includes the sodium pump
(131). Activation of Src requires phosphorylation at the activating Tyr418 site, allowing the
kinase to induce autophosphorylation of neighbouring EGFR (137). Activated EGFR
subsequently recruits SHC (Src homology 2 domain containing), GRB2 (growth factor receptor-
bound protein 2) and SOS (Son of Sevenless), which induces the activation of the RAS-RAF and
ERK1/2 protein cascade (66, 136, 137, 139-141).
23
Signaling of PLC IP3 and intracellular Ca + changes
In parallel to Src activation and in a process not fully understood, PLC and IP3 in a complex with
the Na+/K+-ATPase become activated. This complex of proteins brings the sodium pump in close
proximity to the endoplasmic reticulum, inducing single and repeated transient changes of
intracellular Ca2+ (142).
NF-κB signaling
Activation of NF-κB occurs downstream of Src activation and [Ca2]i oscillations. In addition,
activation of RAS by Src activation leads to the opening of ATP-sensitive mitochondrial
potassium channels, resulting in a concomitant production of reactive oxygen species, triggering
NF-κB (66, 136, 137, 139-141). In parallel to SRC activation, low frequency calcium oscillations
(4-6 minute) in response to low concentrations of ouabain (0.1- 10 nM) also invokes NF-κB to
induce gene programs which leads to differentiation or proliferation (133, 143).
PI3K-Akt signaling
PI3-K is another part of the signaling complex of Na+/K+-ATPase-EGFR and Src. PI3-K is
bound to a proline-rich region of the catalytic α subunit (144). CG binding to the Na+/K+-
ATPase is thought to stimulate PI3-K, and Akt via phosphorylation by PI3-K (145).
1.3.6 CGs: Mechanism of Antitumour Activity
Anticancer effects have been reported in several types of carcinoma, including breast (113, 146,
147), prostate (148, 149), lung (NSCLC) (150-152), melanoma (153), pancreatic (154, 155)
leukemia (156, 157), renal adenocarcinoma (158), colorectal adenocarcinoma (159, 160),
glioblastoma (161), sarcoma (Kaposi sarcoma) (162), osteosarcoma (163), retinoblastoma (164)
and neuroblastoma (128, 139, 165). As described below, CGs induces apoptosis and disrupts
numerous cellular processes vital to tumour growth, such as inhibiting proliferation,
angiogenesis, cell migration and invasion (see Appendix Table 13).
Cell death response to CGs
Cell death is the most commonly reported mechanism of CG anti-cancer effects. In
neuroblastoma cells, namely SH-SY5Y, apoptosis is induced in response to ouabain at 10-100
nM concentrations. In addition, UNBS1450, a novel CG analogue preferentially induces
autophagy in glioblastoma and NSCLC cell lines (98, 152). Oleandrin was also used in NSCLC
24
and sensitizes NSCLC cells to Apo2L/TRAIL induced apoptosis via up-regulation of Death
Receptors 4 and 5. Interestingly, stabilizing DNA topoisomerase II was demonstrated as the
mode of antitumour effects in MCF-7 breast cancer cells treated with 100 nM of digoxin,
ouabain or proscillaridin (166). MCF-7 and other estrogen receptor positive breast cancer cells
may be subjected to dual targeting of the Na+/K+-ATPase and estrogen receptor (ER) by CGs
(167). Reports of computer based alignment analysis predict that the amino acid residues 323-
395 of ERα and likely ERβ are similar to the ouabain binding site of the Na+/K+-ATPase and
could allow for CGs to bind antagonistically to the estrogen receptor (167).
CGs induce differentiation and cell cycle arrest
In contrast to activating cell death, CGs have also been reported to induce differentiation and cell
cycle arrest. Human breast cancer cells MDA-MB-435 (ER negative), following ouabain
treatment (<100 nM) have increased levels of p21 and undergo cell cycle arrest (61). Also,
Bufalin has been shown to induce differentiation of human leukemia cells (HL60, U937 an ML1)
toward a Macrophage/Monocyte fate by cell cycle arrest in the G2-S phase (168-170). In
addition, UNBS1450 is thought to induce cell cycle arrest in prostate cancer cells. Thus far only
three reports have cited that CGs can induce cell cycle arrest or differentiation and there is no
proposed mechanism. Lastly, to date there is no evidence supporting differentiation and cell
cycle arrest in neuroblastoma cells in response to CG treatment.
Antiangiogenic effects of CGs
CGs can also regulate the secretion of FGF-2 (fibroblast growth factor-2) in prostate cancer cells
(PC3 and DU145), which can induce angiogenesis and is involved in a variety of biological
processes (171). The concentrations of ouabain, oleandrin, and Anvirzel used to inhibit FGF-2
release were shown not to inhibit the viability and proliferation of the prostate cancer cell lines,
suggesting that the affect of CGs on angiogensesis was not due to antiproliferative effects (171).
In addition, previously digoxin has been shown in neuroblastoma to have antiangiogenic effects.
Specfically, Svensson et al. demonstrated that treatment of neuro-2a tumours in vivo with
digoxin decreases vascularization induced by FGF-2 in when cells are injected into
chorioallantoic membranes (165).
CGs inhibit the metastatic potential of tumour cells
25
In addition to antiproliferative effects, UNBS1450 (for structure see Fig. 2), impairs migration
and actin cytoskeleton in glioblastoma cells (10 to 100 nM). The anticancer effects of CGs may
also be explained by inhibition of N-glycan expression. Aberrant N-linked glycans promote
tumour invasiveness; consistent with effects on N-glycan cell surface expression, digoxin
inhibited tumour cell migration, invasion and distant tumour formation (172). CGs were also
identified in a high-throughput screen for agents that restore sensitivity of tumour cells to
anoikis. Anoikis occurs as epithelial cells undergo cell death in response to detachment from
cell/cell and cell/extracellular matrix contacts. Cancer cells are thought to be resistant to anoikis
allowing for cell migration and invasion (173). This screen identified five members of the family
of CGs as anoikis sensitizers (ouabain, peruvoside, digoxin, digitoxin, and strophanthidin),
suggesting in addition to the previous described studies that CG anticancer activity may also be
in part due to inhibition of tumour cell migration and invasion.
1.3.7 Cardiac Glycoside: Epidemiological evidence of Anticancer Properties
Stenkvist et al. first reported the potential anticancer effects of CGs in a cohort of women
receiving treatment for breast carcinoma. The study found that patients who were coincidentally
receiving CGs to treat cardiac dysfunction had breast cancer cells that appeared morphologically
more benign, relative to cancer cells from non-CG treated patients (174-176). Furthermore, 5
years following mastectomy, the rate of relapse for patients on digitalis treatment was 9.6 times
lower compared with patients not receiving CG treatment. A 20-year retrospective study on the
same patient group revealed that the death rate from breast carcinoma (excluding confounding
factors) was 6% (two of 32) among patients on digitalis therapy (mainly digoxin or digitioxin),
compared with 34% (48 of 143) among patients not on digitalis (113). Thus clinical evidence
would support further clinical studies into CGs as a potential novel class of anticancer agents.
1.3.8 CGs: Anti-tumour Activity in vivo and in Clinical Studies
Although, there are in vitro and epidemiological studies supporting the anti-cancer properties of
CGs, there is less in vivo evidence supporting these drugs as effective anticancer agents. The first
study fifteen years ago, reported that digitoxin inhibited tumour formation of mouse skin and
mouse pulmonary tumours (177). Recently, anticancer activity of UNBS1450 was demonstrated
26
in xenografts of glioblastoma, prostate cancer and NSCLC (151, 161, 178). Also studies have
shown that digoxin can inhibit cell migration with in vivo models of metastatic prostate cancer
cells (172). In another model for prostate cancer, circulating prostate cancer cells developed
fewer tumours when treated with ouabain relative to non-CG treated controls (173). In addition,
Svensson et al. reported that digoxin induced inhibition of neuroblastoma xenografts.
Interestingly, inhibition of tumour cell growth was targeted against neuroblastoma cells (44%
inhibition of SH-SY5Y cells), whereas lung cancer cells in this study did not respond to digoxin
treatment. This result would suggest that CGs have antitumour potential against only some
tumours, including NB. These studies represent most if not all the in vivo studies to date on the
anticancer effects of CGs. One reason for the lack of in vivo models is the lack of a well-accepted
pre-clinical model. Most models for cancer study use rodent models; however, given the
insensitivity to CGs of most rodents these models are unable to predict toxicity associated with
drug treatment.
Thus far results from clinical trials on CGs are inconclusive as to the effectiveness of these
compounds to treat cancer. One completed phase II clinical trial studied the effects of digoxin in
combination with the EGFR targeting antibody erlontib in NSCLC. The rationale for this clinical
study was based on the association of the Na+/K+-ATPase with the EGFR receptor. No
improvement was detected in the response rate of patients with combined erlontib and digoxin
therapy compared to single agent erlontib (179). More promising results were observed in
treating melanoma patients with digoxin in combinatorial therapy that included cisplatin,
vinblastine, dacarbazine, interleukin-2 and interferon. Preliminary evidence suggests that digoxin
had a synergistic effect on tumours with minimal treatment related toxicities (180). There is one
known documented clinical trial of CGs as a single agent. Anivirzel (a nerium oleander extract
containing predominately oleandrin) was safely administrated to patients in a phase I clinical trial
(0.8 ml/m2/day intramuscular injection). Unfortunately, these patients, all of whom had
refractory disease and were heavily pre-treated did not display signs of tumour regression (181).
The assumption that these drugs may lack antitumour activity at tolerated dose levels possibly
explains why there are only a few clinical trials conducted with these compounds. Although
traditional CGs have not been well-studied in clinical trials, in vitro and in vivo experiments have
shown that novel CG analogues may be effective anticancer agents with predicted diminished
toxicity and for these reasons may be superior to traditional CGs. UNBS1450 is the only known
27
novel CG analogue under phase I clinical trial and is currently being tested for efficacy against
solid tumours and lymphomas in Belgium (182).
1.3.9 Novel Strategies to Circumvent Cardiac Glycoside Induced Toxicity
Stenkvist et al. found that in a retrospective analysis of patients receiving concurrent
chemotherapy and digitalis, therapy resulted in increased survival for breast carcinoma patients.
However, the risk of cardiotoxicity associated with anticancer doses of CGs compounds in part
has stalled further clinical investigation into their anticancer potential. This is consistent with a
narrow therapeutic window for digoxin of 0.6-1.9 nM and digitoxin of 13-33 nM to treat cardiac
dysfunction.
Recent reports of novel CG derivatives suggest that these molecules have increased anti-
proliferative effects with predicted decreases in related toxicity. For example,
neoglycorandomization, the addition of novel glycosides to the C3 carbon of the tetracyclic
steroid ring, showed the importance of the stereochemistry of the C2 carbon in the sugar moiety
of CGs. In particular, a (S) configuration made CG analogues more potent against a panel of
cancer cell lines with concurrent decrease in Na+/K+-ATPase inhibition relative to digitoxin
(183). In addition, the modification of the C19 substituent to a sterically non-hindered polar
group was important in maintaining anti-tumour effects while increasing in vivo tolerance of the
novel cardiac glycoside UNBS1450 (114). In these studies in vivo tolerance was the maximum
dose tolerated by mice that did not cause death over approximately one month(114). Although
mice are not an appropriate preclinical model to assess toxicity, an increase of in vivo tolerance
for UNBS1450 relative to the starting compound would suggest that modifications to the
tetracyclic steroid ring may increase the therapeutic potential of these drugs (114). UNBS1450
also revealed that the Na+/K+-ATPase is a target for the anticancer properties of CGs, but that
inhibition of the Na+/K+-ATPase by CGs do not necessarily correspond to the effects of
antitumour activity (182). Therefore, this would suggest that modifications to the structure of
CGs, in particular the steroid nucleus but also the lactone ring and sugar moiety may improve
anticancer activity without Na+/K+-ATPase inhibition, which is thought to lead to cardiotoxicity.
28
Chapter 2 Rationale, Aims and Hypothesis
2.1 Rationale Neuroblastoma accounts for 15% of paediatric-oncology related deaths (2). Approximately half
of all patients are diagnosed with high-risk NB (2-4). Over the past three decades survival rates
of high-risk patients have not improved despite the intensification and combination of available
therapies (2-4).
In a concerted effort to better understand the biology of NB and develop novel therapies to
increase overall survival, the Kaplan laboratory has identified tumour-initiating cells (TICs)
harvested from the metastases and primary tumours of NB patients. As few as ten NB TICs
injected orthotopically into SCID/Beige mice form neuroblastoma tumours; thus, given their
potent tumour-initiating capacity, NB TICs likely play an integral role in treatment response and
treatment failure. Therefore, drugs that specifically target NB TICs may prove to be effective
against NB.
A high-throughput screen was preformed to identify drugs that selectively target NB TICs
compared to non-transformed control cells with neural crest cell properties termed skin derived
precursors or SKPs (184). Four different chemical libraries, LOPAC, Prestwick, Spectrum and
Biomol were used to test 4687 licensed drugs. The class of CGs was the largest group identified
with 13 compounds showing moderate selectivity against NB TICs (184).
In vitro and in vivo studies would suggest that traditional CGs have anticancer potential against
neuroblastoma. However, a major impediment to the clinical use of traditional CGs is the
cardiotoxicity associated with these drugs. Cardiotoxicity results from inactivation of Na+/K+-
ATPase leading to diminished gradients of Na+ and K+ ions across the cell membrane.
Studies suggest that the potential antitumour activity of CGs can be separated from inhibition of
the Na+/K+-ATPase. To achieve a more potent drug with lower predicted cardiotoxicity, we
tested novel CG compounds synthesized by the Focus in Synthetic Chemistry (FISC) group at
the Hospital for Sick Children for antitumour activity on NB TICS.
29
2.2 Hypothesis We hypothesize that CGs are potential treatment agents for NB and modifications to the
functional groups will identify structural components of CGs and will lead to a compound with
greater antitumour activity and lower predicted cardiotoxicity.
2.3 Project Aims The first aim is to ascertain the effective inhibitory doses of CGs and novel CG analogues in NB
cells. To test the efficacy of parent and novel CG analogues, sphere assays will be conducted on
NB TICs and a metabolic activity assay (MTT assay) on adherent cell lines. A combination of
NB TICs and NB adherent cell lines are being employed because NB TICs are thought to be a
highly relevant model for drug studies and NB adherent cells are more commonly used and
widely accepted in the NB field. As well, immunoblotting for cleaved-PARP will be used to
assess induction of apoptosis following drug treatment.
The second aim is to assess whether the inhibitory activity on the Na+/K+-ATPase is divergent
from potential anti-tumour activity of CGs on NB cells. To characterize the effects of CGs on the
Na+/K+-ATPase, gene silencing and expression of an insensitive form of the Na+/K+-ATPase will
be used with CG treatment. In addition, Na+/K+-ATPase inhibitory activity is thought to play a
predominant role in cardiotoxicity of CGs, thus relative inhibition of the Na+/K+-ATPase in
HEK-293 cells will be used as a surrogate test for potential cardiotoxicity induced by CGs.
The third aim is to identify potential signaling mechanisms involved in the antitumour activity
of CGs. Reports suggest that several signaling proteins are up-regulated downstream of CG
binding to the Na+/K+-ATPase in a Na+/K+-ATPase-mediated signaling transduction pathway.
One such protein is ERK, which has been described to be up-regulated following CG treatment
in neuroblastoma cells and will be the focus of this aim.
30
Chapter 3 Material and Methods
3.1 Cell Systems Investigation of the antitumour activity of CGs on neuroblastoma involved the use of several NB
cell lines including, NB TICs (NB12, NB67, NB88 and NB 122), established adherent NB lines
(IMR-5 and SK-N-(BE-2)) and established NB cell lines that grow as spheres in NB TIC
growing conditions (KCNR) (Table 4). Although preliminary screens used NB TICs there are
several reasons for the incorporation of immortalized adherent cell lines in this project,
including: 1) established adherent cell lines were used to optimize conditions for NB TICs, 2) to
observe any specificity towards NB TICs verses established NB cells, and 3) because established
adherent cells are commonly used and widely accepted in the NB field. Thus testing CGs against
both NB established adherent cells and NB TICs increases the veracity of CGs as effective
therapeutic agents against NB. Lastly, SKPs were used as the non-transformed control. SKPs are
derived from the dermal papilla of foreskin and display neural crest cell properties (185, 186).
Also similar to NB TICs, SKPs grow in serum free conditions and require SKP conditioned
media for optimal growth (185, 186).
Table 4. Characteristics of cell lines
Cell line p53 status MYCN amp Other
Established adherent cell lines
IMR-5 wt amp Adherent cell line KCNR n/a amp Grows as spheres in neural crest stem cell media SK-N-BE-2 C135F amp Adherent Cell line NBTIC lines NB 12 wt non-amp Grows as dense spheroids NB 67 n/a non-amp Grows as dense spheroids NB 88 wt non-amp Grows as lose spheroids NB 122 wt non-amp Semi-adherent SKPs FS90 n/a n/a Grows as dense spheroids FS99 n/a n/a Grows as dense spheroids FS217 n/a n/a Grows as dense spheroids Abbreviations: wt, wild-type; amp, amplification; NA, not available. (187-189).
31
3.2 Cell Culture
NB and SKP cells were incubated in 5% CO2 at 37 °C in culture media listed in Table 5 for each
cell line. NB TICs were isolated and cultured as previously described by Hansford and
colleagues (51). Briefly, cells were isolated from the bone marrow metastasis of high-risk
patients and cultured in vitro with neural crest stem cell media containing 1%
penicillin/streptomycin (100 units/mL). Sphere lines were dissociated to passage cells into a new
flask and for experiments early passages of sphere lines were used (< passage 12).
Table 5. Culturing conditions of cell lines
Cell line Media Supplements
Immortalized neuroblastoma cell line IMR-5, SK-N-BE-2
DMEM (Invitrogen) 0.1 mM non-essential amino-acids, 1mM sodium pyruvate, 10% (FBS)
SKPs FS90 FS99 FS 217
1/2 conditioned media and 1/ 2 fresh media (DMEM: F12; 3:1 (Invitrogen))
Fresh media containing 2% B-27 supplement (Invitrogen), 40 ng/mL bFGF, (Collaborative), 20 ng/mL EGF (Collaborative), 100 U/mL penicillin/ streptomycin
Neuroblastoma Tumour-Initiating Cells NB 12, NB 67 NB 88, NB 122
DMEM: F12; 3:1 (Invitrogen)
2% B-27 supplement (Invitrogen), 40 ng/mL bFGF, (Collaborative), 20 ng/mL EGF (Collaborative), 100 U/mL penicillin/ streptomycin
Abbreviations: bFGF, basic fibroblast growth factor DMEM, Dulbecco’s modified eagle’s medium; FBS, Fetal Bovine Serum; EGF, epidermal growth factor.
3.3 Cell proliferation and viability assays
Alamar Blue Assay
The alamarBlue® assay (Biosource International, Camarillo, CA, USA) indicates viability and
proliferation of cells. Resazurin, a non-fluorescent indicator dye is converted to a red-fluorescent
resorufin via reduction reactions of metabolically active cells. Cells were seeded in a 96-well
plate and treated with drugs in triplicates. At the end of treatment, 10% alamar blue was added to
each of the wells followed by 24 hour incubation in 5% CO2 at 37 °C. Fluorescence was
measured using the FLx 800 fluorescence plate reader (ELx800 Universal Plate Reader, Bio-Tek
32
Instruments) with excitation wavelength of 560 nm and emission wavelength of 590 nm. For
each well the amount of fluorescence is directly proportional to the number of viable cells.
MTT Assay
To assess viability and proliferation MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide] (Roche Applied Science, Laval, QC, Canada), a tertrazolium salt was used and is
reduced in viable cells to purple formazan crystals. BioReduction of MTT occurs mostly by
mitochondrial enzymes primarily succinate dehydrogenase and by NAD(P)H in the cytoplasm
(190). Cells were seeded in a 96-well plate in triplicate. At post-treatment, for adherent cells old
media was aspirated and cells were replenished with 100 µL fresh media containing 10% MTT
reagent. Following, incubation in 5% CO2 at 37 °C for 4 hours, 100 µL of solubilization solution
was added to solubilize formazan crystals. After a second incubation at 37 °C for 24 hours,
soluble formazan was measured at 560 nm absorbency using the ELISA plate reader (ELx800
Universal Plate Reader, Bio-Tek Instruments). Absorbency of formazan directly correlates to the
number of viable cells.
Trypan Blue Assays
Cell viability was measured using the trypan blue exclusion test to confirm results from alamar
blue and MTT assay (Sigma-Aldrich). Non-viable cells stain blue, while viable cells with intact
cell membranes do not absorb the trypan blue dye. Cells were first seeded in 24-well plates and
treated with drugs in triplicate. Non-adherent cells were collected after washing once with PBS.
Cells were centrifuged at 2600 rpm for 5 min. The supernatant was aspirated and 100 µl of 1:1
solution of trypan blue and PBS was added to the cell pellet. Cells were then resuspended,
incubated in the solution for 2 min and counted using a hemocytometer. A viable cell appeared
clear and a non-viable cell blue. At least 100 cells were counted and the results are expressed as a
percentage of viable cells to the total number of cells (viable and non-viable cells).
Sphere Formation Assay
Drug efficacy was tested by sphere formation assay with the help of Mayumi Fujitani and
Kristen Smith. NB TICs (NB12, NB88R2, NB67 and NB122R) and established NB cells
(KCNR), as well as SKP cells were used to test CGs and novel CG analogues. Cell lines were
triturated into single cells and seeded in triplicate into 96-well non-tissue culture treated plates
(Unibioscreen) at 2000 cells per well (NB12, NB88, NB122R) and 3000 cells per well (KCNR,
33
SKPs, NB67) in 50 µl of media, which contained 50% cultured SKP media in experiments
conducted with a SKP line. Upon treatment, 50 µl of media containing twice the indicated drug
concentration was added to each well to 100 µl final volumes. Three days after seeding cells,
wells were retreated with the indicated drug concentration. On day 7, wells were counted for
number of spheres. Results are expressed as percent number of spheres relative to untreated
vehicle controls (0.5% DMSO or H20).
Potassium Chloride Assay
To assess the antagonistic effects of K+ on CG-mediated antitumour activity, NB88R2 were
seeded in 96-well non-tissue culture treated plates with the indicated concentrations of KCl (10
mM, 15 mM and 20 mM). Cells were treated with indicated drug concentration on Day 0 and
Day 3 as described with the sphere formation assay. On day 7, wells were incubated with 10%
alamar blue and viability and proliferation was assessed as described with the alamar blue assay.
3.4 Drug treatment
Chemical suppliers and the stock concentration of novel agents, conventional CGs and
chemotherapeutic agents are listed (Table 4). Novel CG analogues were synthesized by the
Focus in Synthetic Chemistry at the Hospital for Sick Children. Drugs were either dissolved in
dimethylsulfoxide (DMSO) for conventional CGs, H20 for novel CG analogues or in culture
media for doxorubicin. For sphere assays and immunoblots drugs were serially diluted in the
respective control media and further diluted in culture media to the indicated concentration. Final
concentrations of DMSO or H20 in media for drug treatment was less than or equal to 0.05%.
Table 6. Conventional CGs and chemotherapeutic agents used for drug efficacy studies
Abbreviations: Conventional CGs are dissolved in DMSO. Doxorubicin is dissolved in media.
Drug Company Stock concentration
Conventional CGs Convallatoxin Sigma-Aldrich 25 mM Digitoxin Sigma-Aldrich 25 mM Digoxin Sigma-Aldrich 25 mM Ouabain Sigma-Aldrich 25 mM Peruvoside Sigma-Aldrich 25 mM Chemotherapeutic Agents Doxorubicin (DOX) Sigma-Aldrich 1 mg/ml
34
3.5 Na+/K+-ATPase activity screen Na+/K+-ATPase activity was assayed by Aurora Biomed (Vancouver, B.C., Canada). Briefly,
HEK-293 cells, which endogenously express Na+/K+-ATPase, were exposed to rubidium (Rb+)
and subsequently washed with SPA-Wash Buffer and lysed. Cellular uptake of Rb+ in response
to drug treatment was measured using flame atomic absorption spectroscopy (ICR8000, Aurora
Biomed). The Na+/K+-ATPase recognizes Rb+ ions as analogous to K+ ions and thus decreased
Rb+ uptake is considered a marker for inhibition of Na+/K+-ATPase mediated ion influx.
3.6 Immunoblotting (IB) For immunoblotting, spheres lines less than ten passages were dissociated into single cells and
3×106 cells were seeded in 6.7 ml of media in a 25 cm2 culture flask. Upon treatment, 3.3 ml of
drug-containing media at three times the indicated concentration was added to each culture flask
to adjust for dilution. Cells were then pelleted by centrifugation at 4°C and washed with cold
PBS. Subsequently, cells were lysed at 4°C for 30 min using EBC buffer (50 mM Tris [pH 8.0],
120 mM NaCl, 0.5% NP-40) supplemented with protease inhibitors consisting of aprotinin (11.5
µg/mL), leupeptin (1 mg/mL), sodium orthovanadate (0.2 mM), and phenylmethylsulfonyl
fluoride (50 µg/mL PMSF; Roche Applied Science, Laval, QC, Canada). The protein
concentration of whole cell extract (WCE) was measured using the Bradford reagent (Bio-Rad
laboratories, Hercules, CA, USA). Protein was denatured by boiling in SDS-containing sample
buffer at 100°C for 5 min. Equal amounts of protein were resolved by SDS-polyacrylamide gel
electrophoresis (SDS-PAGE). 10% Gels were used to resolve proteins larger than 30 kDa and
15% gels for proteins smaller than 30 kDa. Proteins were transferred to nitrocellulose membrane
at 0.1 Amp overnight at 4°C or 0.4 Amp for 6 hour at 4°C; afterwards, the membrane was
blocked in Tris-buffered saline (TBS) with 5% milk for 1 hour at room temperature (RT) or at
4°C overnight. The membrane was then probed with primary antibody overnight at 4°C, washed
and incubated at RT for 1 hour with HRP-conjugated secondary antibody or infrared dye-tagged
secondary antibodies (Table 7). Nitrocellulose membranes incubated with HRP-conjugated
secondary antibody were washed 5 times with TBS with 0.1% Tween. Protein-antibody complex
was detected using an enhanced chemiluminescence system (Super Signal West Pico, Pierce,
Rockford, Il, USA). For nitrocellulose membranes probed with infrared dye-tagged secondary
antibodies, the protein-antibody complex was detected using the Li-Cor Odyssey infrared
imaging system.
35
3.7 Immunofluorescence (IF) 100 µl of medium containing suspended spheres was removed from a flask and spun down onto
coated slides using a Thermo Shandon Cytospin 4 apparatus (Thermo Shandon Inc., Pittsburgh,
PA,). Afterwards, slides were air-dried for 5 minutes, washed three times with PBS, fixed with
4% paraformaldehyde in PBS for 15-20 minutes and then washed three times with PBS. Cells
were permeabilized with 0.2% NP-40 for 5 minutes, washed three times for 5 minutes with HBS,
and blocked for 1 hour at room temperature with 0.5% bovine serum albumin and 6% normal
goat serum in HBS. Primary antibodies were then added in HBS containing 3% normal goat
serum and left overnight at 4°C. Primary antibody was removed, cells were washed three times
for 5 minutes with HBS, and incubated for 1 hour at room temperature with Alexa 488-
conjugated goat-anti-mouse (1:1,000) or Alexa 555-conjugated goat-anti-rabbit (1:1,000;
Molecular Probes) (Table 7). Cells were washed three times for 5 minutes with HBS, stained
with DAPI, and visualized using a Zeiss Axioplan upright microscope.
3.8 Antibodies
Table 7. List of Antibodies used in immunoblotting, and immunofluorescence studies
Primary Antibody Company Dilution Factor
β actin
Sigma-Aldrich, Oakville, ON, Canada 1/ 30,000
cleaved-PARP Cell Signaling Technology, Beverly, MA, USA 1/750
Na+/K+-ATPase α1 Cell Signaling Technology, Beverly, MA, USA 1/500
Phospho-ERK 1/2 (Thr202/ Tyr204)
Cell Signaling Technology, Beverly, MA, USA 1/1,000
Phospho-Src (Tyr416) Cell Signaling Technology, Beverly, MA, USA 1/1,000
Total-ERK1/2 Cell Signaling Technology, Beverly, MA, USA 1/1,000
Total-Src Cell Signaling Technology, Beverly, MA, USA 1/1,000
Vinculin Upstate, Lake Placid, NY, USA 1/20,000
36
Secondary Antibody Company Dilution Factor
HRP-conjugated anti-mouse IgG (H + L)
Pierce Biotechnology, Rockford, IL, USA 1/5,000
HRP-conjugated anti-rabbit IgG (H + L) Pierce Biotechnology, Rockford, IL, USA 1/5,000
IR Dye-tagged anti-mouse (520 nm) Li-Cor Biosystems, Lincoln, NB, USA 1/5,000
IR Dye-tagged anti-rabbit (620nm) Li-Cor Biosystems, Lincoln, NB, USA) 1/5,000
3.9 Plasmids
The pcDNA3.1-Atp1a1 (murine Na+/K+-ATPase) plasmid was kindly provided by Dr. Schimmer
(Ontario Cancer Institute). The plasmid was amplified by transformation of competent E. coli
bacteria and DNA from several colonies was purified by minipreparation (Qiagen). The plasmid
was sequenced and PCR was performed using murine Atp1a1 specific primers.
3.10 Reverse-transcriptase polymerase chain reaction RNA isolation and RT-PCR
Total RNA was isolated using TRIzol®, a mono-phasic solution of phenol and guanidine
isothiocyanate (Invitrogen), according to manufacturer’s instructions. RNA was then dissolved in
20-30 µl of DEPC treated water and the concentration of RNA was determined by
spectrophotometry. Omniscript RT Kit (Qiagen) was used to generate cDNA; cDNA was then
amplified for semi-quantitative rt-PCR using Taq DNA polymerase (Qiagen). ATP1a1 and β-
actin genes were amplified using the conventional protocol: (1) initial denaturation (95°C/3min),
(2) 32 cycles of denaturation (94°C/1 min), annealing (55°C/ 1min) and extension (72°C/ 1min)
and (3) final extension (72°C/ 10min). Primer sequences are shown in Table 8.
37
Table 8. Sequence of RT-PCR Primers Gene Primer sequence
F 5’- TGT GAT TCT GGC TGA GAA CG -3’ ATP1A1 set #1
R 5’- TCT TGC AGA TGA CCA AGT CG -3’
F 5’- TGGAGTTCACCTGCCATACA -3’ ATP1A1 set #2
R 5’- AGGTTTCCTTCTCCACCCAG -3’
F 5’-ATA AAC CTT GCC CGC TGT CG -3’ Atp1a1 (murine specific)
R 5’- TCC CCT ACT CCC TTC TCA TC -3’ Abbreviations: F, forward; R, reverse.
3.11 Short-interfering RNA (siRNA)
siRNA oligonucleotides sequences were obtained from Dharmacon (Lafayette, CO, USA) and
resuspended according to the manufacturer’s instructions. Cells were transfected with a siRNA
pool of four oligonucleotides. Sequences are shown in Table 9.
Table 9. ON-TARGETplus SMART pool sequences Human ATP1a1, NM_001001586 Target Sequence
ATP1a1 #1 UGA AUU UCC CUA UCG AUA A
ATP1a1 #2 GUU CAA AGA UCA UGG AAU C
ATP1a1 #3 GGU UGG ACG UGA UAA GUA U
ATP1a1 #4 GAA GGC ACC GCA CCG CAC GUG GUA Abbreviations: F, forward; R, reverse
3.13 Transfection of neuroblastoma cells
Three transfection methods were used to insert nucleic acid into NB cells. Electroporation was
utilized to transfect NB88R2 with siRNA. Lipofectamine was used to insert siRNA and PEI to
insert plasmids into IMR-5 cells. Various transfection techniques were used based on
transfection efficiency for each type of nucleic acid and cell type.
Transfection with Electroporation
NB TICs were electroporated with siRNA or plasmid using the pipette-type microporator MP-
38
100 (Digital Bio Seoul, Korea). Briefly, 2 x106 cells were suspended in 100 µl of R buffer
(provided in microporation kit) and mixed with 100 nM of specific or control siRNA
oligonucleotide prior to transfection. To transfect cells the following settings were used: 1350V,
2 ms pulse width and 2 pulses. Samples were then re-counted and seeded into either 6 or 96-well
plates and were incubated for 24 hours in 2 times growth factor media prior to subjecting cells to
any assays. Gene expression was analyzed at 24, 48, 72 and 96 hours following transfection.
Drug treatments were started 24 hours following transfection and cell growth or protein levels
assessed at 72 to 96 hours following transfection.
Transfection with LipofectamineTM 2000
Adherent cells were transfected with oligonucleotide for gene silencing using Lipofectamine
2000 (Invitrogen) using 6-well plates. In one eppendorf tube 3µl Lipofectamine 2000 was
incubated in 250 µl OptiMEM at RT for 20 seconds and in a separate eppendorf tube 3 µl of
DNA was diluted in 250 µl OptiMEM. Lipofectamine was added drop-wise to the diluted DNA
bringing the proportion of DNA to Lipofectamine to a 1 µg to 1 µl ratio. The mixture was gently
mixed and incubated at RT for 20 min. Cells were washed with OptiMEM, and bathed in 2ml/
well with OptiMEM. The DNA/Lipofectamine mixture was added to the cells drop-wise and
gently in a spiral pattern. Four hours later, the contents of each well were aspirated and replaced
with regular culture media and 24 hours later experimental assays were conducted.
Transfection with PEI
Adherent cells were transfected with DNA using PEI in 6-well plates. 3 µg of DNA was diluted
in 400 µl of OptiMEM and PEI (9 µl) was added to the DNA solution to obtain a 3:1 ratio
followed immediately by mixed the solution with a vortex for 10 seconds. The solution was then
incubated for 15 min at RT. The solution was then added and mixed into 6-well plates containing
cells and 2 ml of fresh complete media. The following day, the media was changed and
approximately half a day after changing the media, experimental assays were conducted.
3.12 Statistical Analysis
Data were presented as mean ± standard error of the mean and ‘n’ indicates the number of
independent experiments. For growth assays each independent experiment was performed in
triplicate. Data were analyzed using two-tail paired t-test. Samples were statistically significant
and denoted by * when p < 0.05 and by ** when p < 0.01.
39
Chapter 4 Results
4.1 Digoxin inhibits cell growth of NB TICs and immortalized adherent and non-adherent cell lines
As previously described, CGs were identified in a high-throughput screen for selective inhibitors
of NB TIC growth relative to non-transformed control cells (SKPs) (184). One of the identified
compounds was digoxin and we used digoxin in the following experiments as a CG comparison
for novel CG analogues because it has been previously reported to display anticancer activity and
its therapeutic and toxic effects are well-described (123, 172). Furthermore, we used the sphere
formation assay to identify the effective inhibitory doses of digoxin on NB TICs. We reported
the inhibitory dose as the IC50 value, which is the drug concentration needed to decrease the
overall number of spheres for NB TICS, KCNR and SKPs by 50% following 7 days of drug
treatment. For two immortalized adherent cell lines IMR-5 and SK-N-(BE-2), we report the IC50
value, as the drug concentration required to inhibit 50% growth using the MTT assay, following
3 days of drug treatment. The inhibitory growth curve of cells in response to digoxin treatment is
shown in Figure 5. The IC50 of digoxin on SK-N-(BE-2) and IMR-5 cell growth was 57.4 nM
and 121.3 nM respectively (Fig. 5a). The IC50 of digoxin on NB12, NB88R2 and NB122R was
approximately 37.46 nM, 55.83 nM and 47.37 nM respectively. In addition, digoxin decreased
overall sphere size, suggesting that it also inhibits proliferation of NB TICs, as shown in Figure
5b with NB88R2. The results depicted for NB88R2 are representative for all NB TICs tested and
both sphere size and number decreased in a time dependent manner. For the KCNR cell line the
IC50 was 25.4 nM. For FS99 (non-transformed SKP control) the IC50 was 83.3 nM, suggesting
that digoxin inhibits SKP growth with very little selectivity towards NB TICs (Fig. 5c).
40
Figure 5. Digoxin inhibits neuroblastoma cell growth. A, Cells were treated with various concentrations of digoxin and cell growth of adherent NB cell lines, IMR-5 and SK-N-(BE-2) was assessed by MTT at 72 hours following drug treatment (n=2). NB TICS (NB12, NB88 and NB122R), KCNR and the SKP line FS 99 were treated with increasing concentrations of digoxin and sphere forming potential was assessed at day 7 (B-C).
4.2 Digoxin induces apoptosis in a dose dependent manner in IMR-5 and NB88R2
IMR-5 and NB88R2 were treated with 10 nM, 50 nM and 100 nM of digoxin for 48 hours. Poly
(ADP-ribose) polymerase (PARP), a key substrate of effecter caspases was used to assess cell
death by immunoblotting cell lysates prepared from digoxin-treated cells with an antibody to
cleaved-PARP. Levels of cleaved-PARP directly correlated with an increase in digoxin
concentration in both IMR-5 and NB88R2, suggesting that at least in part digoxin induces
apoptosis to reduce cell viability of neuroblastoma cells (Fig. 6).
41
Figure 6. Digoxin induces apoptosis in a dose dependent manner in IMR-5 and NB88R2. Apoptosis was assessed by immunoblotting for cleaved-PARP 48 hours following drug treatment (n=3).
4.3 Structure-Activity-Relationship Analyses Digoxin has been described to induce cardiotoxicity at concentrations that overlap closely to the
therapeutic doses for treating cardiac dysfunction. The doses that induce cardiotoxicity are also
predicted to be close to effective doses on NB TICs; therefore, in an effort to circumvent
cardiotoxicity we tested several novel CG analogues to identify potential compounds and
structural groups that may contribute to potency (low IC50) and to circumvent predicted
cardiotoxicity. In order to do this we focused on modifications to the three components of a CG:
the lactone ring, the tetracyclic steroid ring and the sugar moiety. The data in Table 10 shows the
structure of compounds synthesized by FISC derived from ouabain, convallatoxin and
peruvoside with their corresponding IC50 values in four NB TIC lines, NB12, NB67, NB88R2,
NB122R; one immortalized NB line, KCNR; and three SKP lines FS 90, FS99 or FS217. The
results highlight the following; (1) the lactone ring is necessary for inhibiting tumour sphere
formation. This was demonstrated with the diminished antitumour efficacy of compounds
containing open lactone rings (RIDK21 and RIDK32) or that exhibited steric hindrance with the
addition of an ethyl derivative (RIDK21, RIDK28 and RIDK32). (2) The sugar moiety may be
sufficient to inhibit sphere formation in TIC lines as displayed with RIDK 10 on NB88R2,
RIDK11 with respect to NB12 and NB122R, and RIDK51 on all tested cell lines. (3)
Modifications to the C19 substituent in the tetracyclic steroid ring affected the potency of CG
analogues. This is evident in the peruvoside-derived (numbered in the 50s) and convallatoxin-
derived compounds (numbered in the 30s) as increasing the negative charge of the C19 bonded
group correlated with higher IC50 values in all cell lines. RIDK34 and RIDK54 were the most
42
potent compounds in their respective subsets (lowest IC50 values). Unlike RIDK54, RIDK34 may
display slight selectivity for NB cells. With the exception of NB67, the IC50 of RIDK34 against
NB cells is 2-3 times lower than the IC50 values for SKP cells (note SKPs n=2). However, this
difference has not been shown to be significant. Alternatively, another compound with greater
selectivity for NB TICs versus SKPs is RIDK52, which relative to NB12 and NB67 is 5 and 25
fold more selective for NB TICs and although promising this result does need to be repeated
(n=1). Thus, on the basis of the observed trend at the C19 bonded position and reports from the
literature that demonstrated the importance of this functional group (see Section 1.3.9), RIDK34
was chosen for further validation as a potential therapeutic agent against neuroblastoma.
43
44
Abbreviations: nd, not determined; --, did not inhibit sphere formation (no effect). The SKP line
tested for a given CG compound is indicated in parentheses.
45
4.4 RIDK34 inhibits cell survival and proliferation of NB TICs
RIDK34 was the most promising compound identified from our structure-activity-relationship
analysis (Table 10). RIDK34 has an oxime function bonded to the C19 carbon, whereas the
starting material convallatoxin has an aldeyde group at this site (Fig. 2). To identify the effective
inhibitory dose (shown also in Table 10), NB12, NB88R2, NB122R, KCNR and SKPs, were
treated with increasing concentrations of RIDK34 and cell growth was assessed by sphere
formation assay. RIDK34 reduced the number of spheres (survival) with IC50 concentrations in
NB12, NB88R2 and NB122R of 8.81 nM, 10.94 nM and 7.43 nM, respectively (Fig. 7a). As
demonstrated with NB88R2 (Fig. 7b), RIDK34 decreased the overall size of NB TICs sphere,
suggesting that proliferation was also inhibited. Lastly, RIDK34 inhibited the growth of the non-
transformed SKP control line FS99 with an IC50 of approximately 25 nM (Fig. 7).
Figure 7. RIDK34 inhibits neuroblastoma cell growth. Cells were treated with increasing concentrations of RIDK34 and sphere formation was assessed on day 7. A & B, sphere formation assay (n=3, except SKPs n=2).
4.5 RIDK34 induces apoptosis in a dose dependent manner in NB88R2
NB88R2 was treated with 5 nM, 10 nM and 20 nM of RIDK34 for 48 hours and apoptosis was
assessed by immunoblotting for cleaved-PARP. RIDK34 induced apoptosis in NB88R2 at low
46
doses (5 nM) and at higher doses of 10nM and 20 nM (Fig. 8). In addition, the concentrations of
RIDK34 that induce apoptosis directly correlate with observed inhibitory doses of RIDK34 in
sphere formation assays, suggesting that at least in part RIDK34 induces apoptosis to reduce cell
growth.
Figure 8. RIDK34 induces apoptosis in NB88R2. Apoptosis was assessed by immunoblotting for cleaved-PARP following 48-hour treatment with RIDK34 at 5 nM, 10 nM and 20 nM (n=3).
4.6 Exogenous expression of murine Na+/K+-ATPase alpha one protects against digoxin-induced inhibition of cell growth in IMR-5
To ask whether CGs exert their antitumour activity through a mechanism that requires the
Na+/K+-ATPase, we expressed the murine alpha one subunit of the Na+/K+-ATPase in NB cells.
Specifically, IMR-5 cells were transfected with pcDNA3.1 containing murine Atp1a1 or
pcDNA3.1 without Atp1a1. Expression of the murine insensitive Na+/K+-ATPase partially
blocked digoxin-mediated growth inhibition of IMR-5 cells compared to cells transfected with
control plasmid (Fig. 9). These results suggest that the antitumour activity of digoxin is in part
mediated by the Na+/K+-ATPase.
47
Figure 9. Expression of CG resistant murine Atp1a1 protects IMR-5 from digoxin-induced inhibition of cell growth. Overexpression was assessed by western blot using the Na+/K+-ATPase antibody (A) and by rt-PCR using murine Atp1a1 specific primers (B). C, Viability and proliferation was assessed by MTT at 24 hours following drug treatment (n=2).
4.7 Digoxin inhibits NB TIC growth independently of its effects on [Na+]i and [K+]i
Studies on digoxin and other CGs show that these compounds abolish transmembrane gradients
of Na+ and K+ via inactivation of the catalytic alpha isoforms of the Na+/K+-ATPase. Thus to
further elucidate the mechanism of CG-induced antitumour activity, increasing concentrations of
KCl was used to study whether digoxin and likely other CGs affect cell growth through
modulation of intracellular Na+ and K+ concentrations. For this experiment, cell growth of
NB88R2 was observed in the presence of increasing concentrations of KCl and digoxin
treatment. Unlike, normal physiological concentrations of extracellular K+ (approximately 5
mM), elevated concentrations of KCl (>5 nM) are able to antagonize the inhibitory effects of
CGs on Na+ and K+ gradients that occur as result of Na+/K+-ATPase inactivation (60). As shown
in Figure 10, elevated concentrations of KCl (10 nM, 15 nM and 20 nM KCL) did not restore
cell growth of digoxin treated NB88R2 cells. This result supports the theory that digoxin-induced
inhibition of cell growth is not directly caused by deregulation of Na+ and K+ gradients. The
findings from Figure 9 and 10 suggest that the anticancer effects of digoxin are in part mediated
by the Na+/K+-ATPase, yet in a manner that does not necessarily require inactivation of the
Na+/K+-ATPase, and diminished Na+ and K+ gradients.
48
Figure 10. High K+ medium does not protect against digoxin-induced inhibition of cell growth. Cells were treated with digoxin in media containing normal physiological concentrations of KCl (5 mM) or increasing concentrations of KCl (10 nM, 15 nM and 20 mM KCl). On the seventh day, cell growth was assessed by Alamar Blue (n=3).
4.8 Na+/K+-ATPase alpha one subunit knockdown sensitizes IMR-5 and NB88R2 to digoxin growth inhibitory effects
NB TICs highly express the α1 subunit of the Na+/K+-ATPase relative to α2, α3 and α4 subunits
(see Appendix Fig. 17). Also, to determine the effect of the Na+/K+-ATPase in CG-induced
inhibition of cell growth, we also silenced the expression of the α1 subunit of the Na+/K+-
ATPase in NB cells (IMR-5 and NB88R2). NB88R2 and IMR-5 cells were transfected with a
siRNA pool targeting ATP1A1 transcripts (alpha one isoform of the human Na+/K+-ATPase) or
with non-targeting siRNA. The siRNA to the Na+/K+-ATPase effectively reduced the expression
of this gene (Fig. 11a and 11c). Silencing the expression of the Na+/K+-ATPase in the presence
of digoxin treatment effectively increased the magnitude of digoxin-induced inhibition of cell
growth in both IMR-5 and NB88R2 (Fig. 11b and 11d). This result suggests two possibilities; (1)
the α1 subunit is required for survival such that drug binding and gene silencing have a
concomitant effect; or (2) the α1 subunit sequesters CGs preventing the therapeutic agent from
interacting with cell death effectors whereby in non-transfected cells, endogenous expression of
the α1 subunit attenuates cell death from CG-treatment.
49
Figure 11. Knockdown of the alpha one subunit of the Na+/K+-ATPase increases the magnitude of digoxin-induced inhibition of cell growth. Knockdown was assessed by western blot and rt-PCR in IMR-5 and NB88R2 respectively (A & C). Viability and proliferation was assessed following drug treatment by either MTT at 36 hours for IMR-5 (n=2) or by alamar blue at 72 hours for NB88R2 (n=3) (B & D respectively). Data was analyzed by Student’s t-test and statistical significance of between RIDK34 treated siControl and siATP1a1 is indicated by * (p<0.05).
4.9 Na+/K+-ATPase alpha one subunit knockdown sensitizes NB88R2 to RIDK34 growth inhibitory effects
NB88R2 was transfected with siRNA targeting ATP1A1 and treated with RIDK34 as previously
described for digoxin. Similar to results with digoxin treatment, knockdown of α1 sensitizes
NB88R2 to RIDK34 growth inhibitory effects, suggesting that α1 is a target for the antitumour
activity of RIDK34 in NB TICs (Fig. 12).
50
Figure 12. Knockdown of the alpha one subunit of the Na+/K+-ATPase increases susceptibility to RIDK34-induced inhibition of cell growth. Knockdown was assessed by western blot, rt-PCR (Fig. 11) and immunofluorescence in NB88R2 (A). Viability and proliferation was assessed as previously described in Figure 11 (B). Data was analyzed by Student’s t-test and statistical significance between RIDK34 treated siControl and siATP1a1 is indicated by * (p<0.05) and ** (p<0.01), (n=3).
4.10 RIDK34, Digitoxin and Convallatoxin inhibit the Na+/K+-ATPase at concentrations higher than the inhibitory concentrations for NB cell growth
Na+/K+-ATPase activity was tested by Aurora Biomed. The Na+/K+-ATPase recognizes Rb+ ions
as analogous to K+ ions and decreased Rb+ uptake is inferred as inhibition of Na+/K+-ATPase
activity. Moreover, inhibition of the Na+/K+-ATPase mediated Na+ and K+ ion flux is associated
with toxic effects, including cardiotoxicity; thus, Na+/K+-ATPase inhibition was used as a
predictor of CG-induced toxicity. The IC50 for Na+/K+-ATPase inhibition was 1.038 µM, 0.749
µM and 1.042 µM for digitoxin, convallatoxin and RIDK34 respectively (Fig. 13). RIDK34,
relative to convallatoxin and digitoxin has the lowest IC50 on NB TIC cell growth yet the highest
IC50 for pump inhibition (Table 11). As a result, RIDK34 is predicted to be more effective and
better tolerated than digitoxin and convallatoxin. From the Na+/K+-ATPase activity assay, we
predict that serum concentrations of RIDK34 that invoke toxicity are similar to digitoxin
(approximately 33 nM), which is above the inhibitory concentration for RIDK34 on NB TIC cell
51
growth (approximately 8 nM). This compound may therefore have a wider therapeutic window
than digitoxin.
Figure 13. Curve fits, and IC50 determination of Na+/K+-ATPase activity inhibition in HEK-293 cells. Aurora biomed tested for Na+/K+-ATPase inhibition using the Rb+ uptake assay for (A), Convallatoxin; (B), RIDK34; and (C), digitoxin.
Table 11. IC50 values for inhibition of NBTIC growth, Na+/K+-ATPase activity and CG serum toxicity.
Drug NB TIC cell growth (IC50)
Na+/K
+-ATPase
inhibition (IC50) Serum concentration for threshold toxicity
RIDK 34 8.16 nM 1.042 uM (HEK-293) nd
Convallatoxin 12.23 nM 0.749 uM (HEK-293) nd
Digoxin 47 nM nd 0.6 to 1.9 nM (125) or
2.15-3.16nM (123)
Digitoxin 300 nM 1.038 uM (HEK-293) 13 to 33 nM (125)
Abberviation: nd, not determined.
52
4.11 Digoxin increases phosphorylation status of Src and ERK1/2
The previous data indicates that in neuroblastoma cells the effects of digoxin and RIDK34 are in
part mediated by the Na+/K+-ATPase (Fig. 9, 10 and 11). In addition, studies have demonstrated
that digoxin, ouabain and other CGs activate a Na+/K+-ATPase-mediated signal transduction
pathway involving Src kinase and ERK1/2 activation (129, 130, 138, 191). Kulikov et al.
demonstrated in response to ouabain, ERK1/2 and programmed cell death are activated in SH-
SY5Y NB cells (128), suggesting that the Na+/K+-ATPase-mediated signal pathway is activated
in CG treated NB cells. To determine whether Src and ERK1/2 are activated in NB88R2 cells in
response to CG-induced cell death, cell lysates were prepared from digoxin or vehicle treated
cells and probed in western blots with activation-specific antibodies for Src (pSrc Y416) and
ERK 1/2 (pERK T202/ Y204). Our results suggest that Src activity increased relative to vehicle
at 50 nM of digoxin, while ERK 1/2 increased at 50 nM and 100 nM of digoxin (Fig. 14).
Figure 14. Digoxin induces Src and ERK1/2 activation. Cells were treated as described in Figure 2 with digoxin and Src as well as ERK phosphorylation was assessed via western blot using the appropriate antibody. The lower band in Src blots is non-specific for the protein (192, 193). For ERK blots, the upper band is ERK1 and the lower band is ERK2 (n=3).
53
4.12 RIDK34 increases activation of ERK1/2 and at low doses Src kinase
The activity of Src and ERK1/2 was assessed in NB TICs after treatment of cells with RIDK34
(Fig. 15). Src activation was observed only at low doses of RIDK34 treatment (5 nM and 10
nM), likely due to the lack of protein due to cell loss at 20nM. An increase in ERK1/2 was
observed at 10nM RIDK34, with a much larger increase at 20 nM.
Figure 15. RIDK34 induces Src and ERK1/2 activation. Cells were treated with RIDK34 as described in Figure 14. Src kinase and ERK1/2 phoshporylation status was assessed via western blot using the appropriate antibody following treatment with RIDK34 (n=3). The lower band for the Src blot is non-specific for the protein (n=2). For ERK blots, the upper band is ERK1 and the lower band ERK2 (n=3).
54
Chapter 5 Discussion
Neuroblastoma is an enigmatic malignancy with a broad spectrum of clinical outcomes. In
general, current available therapies for high-risk patients with metastatic neuroblastoma are
ineffective and survival rates for these patients are less than 35% (4, 5). Despite intensification
and combinatorial therapies to improve outcome, survival rates have changed very little over the
past three decades for high-risk patients.
One likely reason for the lack of effective therapies is the lack of appropriate in vitro and in vivo
NB models. In an effort to better understand and develop novel therapies to treat NB, NB TICs
were isolated from patients samples. Consistent with tumour-initiating and cancer stem cell
models, NB TICs share many properties with normal stem cells and are capable of self-renewal
and differentiation into the cell types observed in NB (51). Also, NB TICs have enhanced
tumour-initiating capacity and form tumours with as few as ten cells in SCID/Beige mice (51).
Lastly, NB TICs are thought to play an integral role in treatment response, suggesting that
therapies targeting NB TICs are likely to be effective against NB.
In an attempt to identity novel therapies to better treat NB, a high-throughput screen was
performed to identify bioactive molecules that selectively inhibit tumour growth potential of NB
TICs. Of the over 4,000 compounds tested, inhibitors of the Na+/K+-ATPase were the most
commonly identified class of drugs (184). From the 13 CGs identified, 9 were cardenolides and 4
were bufadienolides. Two cardenolides: digoxin and ouabain were previously identified as
potential anticancer agents against NB (128 &164), however, the drug screen completed in the
laboratory of David Kaplan is the first to implicate a large number of CGs as potential agents that
could be used for the treatment of NB.
5.1 Digoxin and the novel CG analogue RIDK34 inhibit growth of NB cells
Traditionally prescribed for the treatment of cardiac dysfunction, CGs also display antitumour
activity in several in vitro tumour models including NB (128). We demonstrated that CGs have
anticancer activity in vitro against immortalized adherent cells SK-N-(BE-2) and IMR-5,
immortalized non-adherent cells KCNR and NB TICs. From modifications of the three parts that
constitute a CG; the lactone ring, sugar moiety and tetracyclic steroid ring, we were able to
55
characterize the necessary functional groups in CG compounds for antitumour activity. As a
result, the potency of CGs against NB was enhanced as evident with our lead compound
RIDK34, which unlike conventional CGs displays anticancer effects at doses below 10 nM (see
Appendix Fig. 19 for reaction scheme).
5.2 Activity of CG-derived sugar groups In testing the effects of CGs and their derivatives on NB TICs, we observed that a few isolated
sugar groups derived from CGs displayed antitumour activity. This was demonstrated with
RIDK10 on NB88R2, RIDK11 with respect to NB12 and NB122R. RIDK 51, the L-thevetose
sugar moiety of peruvoside, appeared to have antitumour activity on all tested TIC lines (Table
10). RIDK10 and RIDK11 are identical in structure and are the rhamnose sugar of ouabain.
RIDK11 is derived from ouabain, while RIDK10 is a pure sugar compound purchased from
Sigma-Aldrich. Interestingly, both these compounds displayed partial antitumour activity; thus,
we can conclude that the remaining ouabain unconsumed in the production of RIDK11 did not
account for the antitumor effects of these drugs. Very little is known about the effects of sugar
compounds on tumour growth and whether the effective micromolar doses on NB TICS in vitro
could be achieved in patients. Given that these CG analogues lack the lactone and tetracyclic
steroid rings necessary for binding to the Na+/K+-ATPase, we expect that these compounds work
through a novel mechanism. Unlike conventional CGs these sugar groups may not be restricted
to the nanomolar range for therapeutic effects. One possible mechanism for the antitumour
activity of RIDK51 is cytolysis resulting from hyperosmotic media incurred from enriched sugar
concentrations. Although we did not test this prediction, cell culture media is typically
supplemented with 5.5 mM glucose and the concentration of RIDK51 required to induce
antitumour activity in NB TICs is 200-250 µM. To achieve this concentration of RIDK51 would
require an increase of the concentration of sugar compounds in the media to 5.75 mM. This
small increase in sugar concentration would likely not affect cell viability, suggesting that the
antitumour response to RIDK51 is not due to hyperosmotic conditions.
Digitoxin compared to its aglycone digitoxigenin (sugar moiety is replaced by a hydroxyl group
in digitoxigenin) exhibits higher CG binding affinity to the Na+/K+-ATPase, which does not
correspond to an increase inhibitory effects on the Na+/K+-ATPase (see Table 3). Also,
Langenhan et al. demonstrated that modifications to the sugar moiety could increase the potency
of CG antitumour activity with no change or reduced inhibition of Na+/K+-ATPase activity.
56
These observations were the basis for the formulation of compounds such as RIDK27, which we
hypothesized has a diminished inhibitory effect on Na+/K+-ATPase activity. Unfortunately, we
did not test compounds such as RIKDK27 for inhibition of Na+/K+-ATPase activity or binding
affinity using competitive [3H] ouabain binding assays. Thus, future work may include structure-
activity assays to identify modifications on the sugar moiety that increase antitumour activity,
binding affinity and decrease inhibition of Na+/K+-ATPase activity. Following this analysis, the
results could be used to create a RIDK34 derivative that displays a wider therapeutic window.
5.3 The C19 substituent affects CG potency
CG analogues demonstrated an increase in antitumour activity corresponding to C19 bonded
groups that were sterically non-hindered and polar. This was exhibited with the antitumour
activity of RIDK52, RIDK53 and RIDK54. The groups bonded to the C19 carbon were an oxime
group for RIDK54, a hydroxyl group for RIDK53 and an amine group for RIDK52. The rank of
antitumour activity in NB12 was RIDK54> RIDK53> RIDK52. Of the three functional groups,
the oxime group has the greatest number of hydrogen bond acceptors and donors and was least
sterically hindered by the tetracyclic steroid ring followed by the hydroxyl group and lastly the
amine group. The convallatoxin-derived compounds RIDK 33, 35 and 34 have identical
modifications to RIDK52, RIDK53 and RIDK54 respectively. These compounds also
demonstrated that a sterically non-hindered polar group bonded to the C19 carbon increases
antitumour activity of CGs.
The contribution of the C19 bonded group to CG antitumour activity is poorly understood, yet
the potency of RIDK34 and RIDK54 may in part be due to increased binding affinity to the
Na+/K+-ATPase. The amino acids glutamine-118 (M1 helix) and asparagine-129 (M2 helix)
appear to be critical to CG binding of the Na+/K+-ATPase because the substitution of these two
residues with arginine and aspartic acid respectively, causes a substantial reduction in CG
inhibition of the rodent α1 subunit (116). The crystal structure of bound ouabain to the Na+/K+-
ATPase, predicts that these two residues potentially form hydrogen bonds with the C19 bonded
groups (116). This model may also support the increased potential of the oxime group with a
high polar character to form stronger and/or a greater number of hydrogen bonds with the at
glutamine-118, asparagine-129 and possibly other residues in the M1-M2 helix.
Alternatively, quantitative-structure-activity-relationship analysis suggest that a hydroxyl group
at the C19 carbon, reduces binding affinity because it diminishes hydrophobic interactions
57
between the β face of CGs (see Fig. 2) and complementary residues on the Na+/K+-ATPase
(119). Although a C19 bonded oxime group would be predicted to diminish hydrophobic
interactions, the added distance of hydrogen bond acceptors and donors from the steroid nucleus
may diminish negative effects on hydrophobic interactions of the oxime group. The structure-
activity-relationship analysis that lead to the identification of UNBS1450, suggests that a
sterically non-hindered polar group, in the case of UNBS1450 a hydroxyl group attached to the
C19 carbon is thought to increase its maximum tolerated dose in rats without decreasing its
potency against tumour cells of NSCLC, GBM and prostate cancer. This would suggest that
RIDK34 with a similar C19 bonded group may be well tolerated and may also exhibit efficacy
against multiple types of cancer.
5.4 The lactone moiety is essential for the antitumour activity of CGs Modifications to the lactone moiety of the starting compounds ouabain, peruvoside and
convallatoxin negated the antitumour effects of the precursor molecule. This was observed with
the open lactones RIDK21 and RIDK32 or with steric hindrance by the addition of an ethyl
derivative in RIDK21, RIDK28 and RIDK32. Our results suggest that the lactone moiety is
necessary for antitumour activity and is consistent with reports showing decreased activity of
CGs with open-ring lactones (118). However, modifications to the lactone ring may still yield
analogues with antitumour activity. The most promising compounds would be CG analogues that
maintain binding affinity yet display diminished Na+/K+-ATPase inhibition. Reports suggest that
CGs with 5-membered lactone rings have higher binding affinity and lower Na+/K+-ATPase
inhibition compared to a 6-membered lactone ring (119). Possibly, CGs with a 4-membered
lactone ring or another slight lactone ring modification will enhance the discrepancy between
binding affinity and inhibition of the Na+/K+-ATPase. Such a molecule would be predicted to
have a wider therapeutic window and to be a useful therapeutic agent against NB (119).
5.5 Mechanism of CG antitumour activity and the role of the Na+/K+-ATPase
To assess the mechanism of CG-induced antitumour activity on NB cells we choose digoxin
because it displayed potency against NB TIC growth and because it is the most widely reported
CG in the literature with well described therapeutic and toxic effects. In sphere formation assays,
we saw a fewer number of spheres with digoxin treatment suggesting that it inhibits cell survival.
Also CGs are widely reported to induce apoptosis (see Appendix Table 13), thus we treated
58
IMR-5 and NB88R2 with digoxin and observed induction of cleaved-PARP in a dose dependent
manner.
To study the role of the Na+/K+-ATPase in CG antitumour activity we used increasing KCl
concentrations, expression of murine Na+/K+-ATPaseα1 and siATP1A1 to silence the expression
of the Na+/K+-ATPase α1 subunit. When expressing the CG-insensitive murine Na+/K+-
ATPaseα1 we observed partial protection against digoxin relative to transfection control. Also
with siATP1a1 we saw an increase in the magnitude of digoxin-induced inhibition of cell growth
relative to siControl (see Section 5.6 for more explanation). These two assays suggested that in
NB cells the Na+/K+-ATPase is a target for CG antitumour activity. Lastly, to assess the role of
the Na+/K+-ATPase in CG antitumour activity, we used increasing KCl concentrations. The
literature reports that >5 mM KCl, partially reduces the binding and inhibition of CGs to the
Na+/K+-ATPase (60, 194). However, increasing [K+] had no affect on the antitumour activity of
digoxin (Fig. 10). To explain this, studies by Orlov et al. showed [K+] lower than 5 mM and
ouabain in 5 mM [K+] both increase [Na+]i yet that only ouabain and not low extracellular [K+]
induces apoptosis. Interestingly, ouabain treatment in low [K+] induces cell death without any
changes in [Na+]i relative to untreated low [K+] control, suggesting that changes in intracellular
Na+ and K+ gradients downstream of Na+/K+-ATPase inactivation is not necessary for
antitumour effects (195, 196). In our assay, we did not test for changes in intracellular gradients,
yet we would expect that increased extracellular K+ would maintain low levels of [Na+]i in the
presence of CG drugs. In addition, recent reports on NB cells show that intracellular [Ca2+],
[Na+] and [K+] are not perturbed by ouabain at concentrations that we show inhibit cell growth of
NB TICs (128), which would further support a mechanism of action that may not require
Na+/K+-ATPase inactivation (Fig. 17).
Although we suggest a mechanism of CG antitumour activity that does not require Na+/K+-
ATPase inactivation, it is still likely that in parallel, Na+/K+-ATPase inactivation affects cell
viability. We propose that the antitumour activity in NB cells in part requires interaction with the
Na+/K+-ATPase, but may not directly require inactivation of Na+/K+-ATPase function and
subsequent deregulation of [Na+]i and [K+]i. This is in part supported by the observation of the
binding affinity of some CGs to the Na+/K+-ATPase, which do not appear to directly correspond
with Na+/K+-ATPase inhibition, suggesting the possibility of disparate affects resulting from CG
binding to the Na+/K+-ATPase. Furthermore, Xie et al. previously reported that Na+/K+-ATPase
59
mutants that do not regulate ion flux, still respond to CGs by activating a signaling cascade that
included Src and ERK (136, 137). Therefore, we looked to see if Src and ERK activity increases
with digoxin concentration and found that both ERK and Src kinase overall appear to increase
with digoxin treatment. These combined results implicate that the activation of a signaling
pathway, involving Src and ERK, is associated with cell death in response to targeting the
Na+/K+-ATPase, possibly independent of inactivation of Na+/K+-ATPase activity.
Figure 16. CGs possibly activate two parallel death pathways in NB TICs. CGs activate an ionic pathway that require Na+/K+-ATPase inactivation and deregulation of ionic gradients resulting in cell death. Also a second pathway called the Na+/K+-ATPase-mediated signaling pathway causes cell death in a manner possibly independent of Na+/K+-ATPase inactivation. Expression of murine insensitive ATP1a1 and silencing endogenous ATP1A1 likely affect both of these pathways because they predominately alter CG-target binding. Although the KCl assay likely affects the ionic pathway to offset deregulation of ion gradients, it is less likely to prevent the activation of a Na+/K+-ATPase-mediated signaling pathway that also induces cell death. The Na+/K+-ATPase-mediated signaling pathway may involve ERK and Src signaling, yet the role of these signaling targets in CG antitumour activity is still poorly understood.
5.6 Silencing of ATP1A1 increases susceptibility to digoxin and RIDK34 antitumour activity
Decreasing the expression of Na+/K+-ATPaseα1 in NB88R2 increased the magnitude of digoxin
and RIDK34 antitumour activity and emphasized the role of the α1 subunit in sustaining viability
and proliferation of NB TICs. Given its essential role in cell physiology, cells likely compensate
for reduced ATP1A1 by trafficking formed oligomers of the pump from intracellular membranes
60
to the plasma membrane and decreasing degradation of the ATPase. Also the redundant roles of
the α1-α4 subunits likely offset the knockdown of ATP1A1.
Presumably, silencing ATP1A1 reduces the overall plasma membrane pool of the Na+/K+-
ATPase without inducing cell death; however, ATP1A1 silencing likely increased cell
susceptibility to further insults targeting the Na+/K+-ATPase explaining increased antitumour
activity in response to digoxin and RIDK34. An alternative interpretation purports that
decreasing the levels of the α1 subunit frees up digoxin and RIDK34 to bind to a greater
proportion of α3 and α4 subunits or an unknown target that more effectively propagates the
antitumour effects of CGs. Evidence suggests that α3 is an important target of CG-induced
antitumour activity, and was demonstrated in cells with high α3 expression, yet given that NB
TICS weakly express α3 and α4 this may suggest a diminutive role of these α subunits in NB
TICs. Alternatively, there may be cellular targets of CGs other than the α subunits of the Na+/K+-
ATPase. Zhang et al. suggested that HIF-1α may represent an additional target of CGs. CGs are
thought to inhibit HIF-1α dependent gene transcription, yet it is unclear whether this function of
CGs may be in part mediated by the Na+/K+-ATPase (170). Additionally, like other steroid
molecules such as estrogen, CGs may cross the plasma membrane and act directly on steroid
receptors in the nucleus to affect tumour cell growth; again this mechanism is still poorly
understood and requires further testing (167). Although we cannot dismiss the role of the α3 and
α4 subunits as well as an alternative target, the results presented here with the expression of
murine Na+/K+-ATPaseα1 and siATP1A1 experiments do indicate that the α1 Na+/K+-ATPase is
a target for CG antitumour activity.
5.7 RIDK34 is a novel CG analogue with higher potency and displays lower relative inhibition of Na+/K+-ATPase activity
We observed that RIDK34 in a series of convallatoxin-derived compounds displayed the greatest
potency against NB TICs. We suggest that this is due to an increase in favourable binding
interactions with its cellular target(s). We also demonstrated that silencing the expression of the
α1 subunit of the Na+/K+-ATPase resulted in an increase in the magnitude of cell growth
inhibition from RIDK34 treatment. Similar results were observed when silencing the expression
of α1 subunit with digoxin treatment. Thus, as with digoxin we believe that this is because the
Na+/K+-ATPase is a target for RIDK34.
61
We observed with RIDK34 treatment a decrease in the total number of spheres with increasing
RIDK34 concentration, thus as with digoxin, we assessed cleaved-PARP to observe if apoptosis
was the mechanism of RIDK34 growth inhibition. We found that cleaved-PARP was activated
by RIDK34 in a dose-dependent manner. Also, the dose of cleaved-PARP induction parallels the
dose of growth inhibitory effects of RIDK34 suggesting that in part RIDK34 induces apoptosis
to reduce cell growth.
Manunta et al. suggested that CG activity (to induce hypertension) does not strictly correlate
with Na+/K+-ATPase inhibition (197). This may be true in our model whereby increasing the
possible number of interactions to the M1-M2 region of the Na+/K+-ATPase may have little
impact on K+ binding (and inhibition) that likely occurs when the lactone ring of CGs bind to the
M4-M6 regions. Thus, the C19 bonded oxime group may increase binding affinity and the
magnitude of cell death through its interactions with the M1-M2 helix without a corresponding
increase in Na+/K+-ATPase inhibition. This prediction is supported by comparing growth
inhibition and Na+/K+-ATPase inhibition of RIDK34 and its starting material convallatoxin, as
the doses of growth inhibition of RIDK34 are slightly lower than convallatoxin for NB TICs
(Table 11), yet the doses for Na+/K+-ATPase activity inhibition is slightly higher for RIDK34
(Table11). This would suggest as predicted by Manunta that inhibition of Na+/K+-ATPase ion
pumping does not directly impact activity of CGs, namely antitumour activity of RIDK34;
however, the two may still be closely related as binding to the Na+/K+-ATPase may impact both
CG antitumour activity and Na+/K+-ATPase inhibition.
Suggesting that RIDK34 has a higher potency and lower Na+/K+-ATPase activity inhibition
implies that in vivo it will have a wider therapeutic window. This would include reduced adverse
effects on the gastrointestinal tract that would normally lead to nausea and vomiting as wells as
reduced vision impairments normally associated with CG treatment such as decreased sharpness
of vision and decreased colour perception (110-112). In addition, inhibiting the Na+/K+-ATPase
leads to Ca2+ overload and ectopic beats, which can cause ventricular arrhythmias and may result
in death if left untreated (202). Thus with RIDK34, we may have slightly decreased Na+/K+-
ATPase activity inhibition and increased antitumour activity. Lastly, if we compare toxic serum
doses in humans of RIDK34 with digitoxin, we find that the effective concentration of RIDK34
(8 nM) is below this range (13-33 nM). Thus, these results would justify further preclinical
studies to more accurately determine the toxicity of RIDK34 with respect to cardiac function.
62
The role of Src and ERK kinase in CG-induced cell death
Our results suggest that the activation of a death effecter pathway occurs following CG-targeting
of the Na+/K+-ATPase. One such pathway could be the putative Na+/K+-ATPase-mediated
signaling transduction pathway involving both Src and ERK, which when activated promote
proliferation and survival or cell death depending on the specific CG, the type of cell and the
effective CG concentration (130-138). In regards to cell death, Newman et al. demonstrated that
pancreatic cancer cells treated with oleandrin have high levels of activated ERK associated with
cell death and when ERK activation was abolished with a MEK1 inhibitor or a specific peptide
inhibitor, cells were partially protected from oleandrin-induced cell death (154). In addition,
studies in NB cells demonstrated that CGs upregulate ERK activity in possibly an α isoform
specific manner as demonstrated by selective knockdown of the α1 and α3 isoforms (128). This
may suggest an integral role for ERK in CG-induced cell death of NB TICs. This prediction is
supported by our observations of ERK activation following digoxin and RIDK34 treatment of
NB88R2. In general, ERK activation is important in cell survival, proliferation and
differentiation; thus, our results generally conflict with published data regarding the role of ERK
on cell viability and proliferation. Interestingly, unpublished data using chemical inhibitors on
NB TICs demonstrated that in a normal setting the survival and proliferation of NB TICs does
not depend on the activity of ERK, JNK or p38 MAPK. Regardless, whether ERK signaling is
necessary for CG-induced cell death of NB TICs is not known and needs to be thoroughly tested
(see Chapter 6).
Also in response to CG-treatment we assessed Src kinase activation and generally observed an
increase in the levels of activated Src, which is consistent with the Na+/K+-ATPase-mediated
signaling hypothesis. This finding is supported with reports showing abrogation of ouabain-
induced ERK activation and neuronal cell death in cell cultures treated with the PP2 Src family
kinase inhibitor (198). However, there are contrary reports that have suggested that Src
activation is not necessary for CG-induced cell death and that the role of Src may be different
depending on mostly the cell type, but also the CG tested and the CG concentration (194, 196).
Generally, Src is thought to promote proliferation in cancer cells and unpublished data
demonstrated that inhibition of Src activity by Src family inhibitors or knockdown reduces
survival and proliferation of NB TICs. This may suggest that Src itself is not crucial for CG-
induced cell death of NB TICs. If ERK activation is involved in CG-induced cell death we could
63
expect an alternative mode of ERK activation. One possible mechanism is through an increase in
ROS production which has been previously described in NB cells in response to CG treatment
(refer to Chapter 6 for experiments to test this prediction) (128).
5.8 Putative role of the α3 and α4 subunit in CG-induced antitumour activity
Despite weak expression of the α3 subunit, this form of the Na+/K+-ATPase may still partially be
involved in the growth inhibitory response of NB TICs to CGs. Recently, Hauck et al. showed
that the binding to human isozyme combinations of α1/β1, α2/β1 and α3/β1 was CG specific
(200). In physiologically relevant conditions of K+, digoxin showed higher binding affinity for
α2/β1 and α3/β1 than for α1/β1 isozymes (200). Thus isozymes more sensitive than α1/β1 may
respond better to CGs to induce antitumour effects. Also weakly expressed α subunits may
respond more effectively to CGs than higher abundance α subunits possibly through localization
in the plasma membrane in caveolar pits (or other key regions) and/or association with important
protein signaling complexes (130-137). This is suggested in SK-N-AS cells, in which ERK
signaling was shown to require only α3 and not the α1 subunit (128). In addition to low α3
expression, NB TICs express low levels of the α4 subunit (see Appendix Fig. 17). Temporal-
spatial expression of α4 in the nervous system has not been described before. Normally
expressed during spermatogenesis in spermatocytes, spermatids and spermatozoa, α4 localizes to
the mid-region of the flagella and its selective inhibition reduces motility as well as intracellular
Ca2+, and pH (65, 81, 199). It is possible that α4 also plays an important role in NB TICs based
on its unusual expression; however, what that role may be and the role of the α3 subunit in CG
inhibition of NB TICs is not known.
5.9 Therapeutic rational for CGs in combinatorial therapy
Retinoic acid (RA) is known to induce terminal differentiation of NB cells in vitro (2). In
addition, after successful phase I and II clinical trials the retinoid, Fenretinide is currently being
formulated for oral administration to young children (2). Similar to its effects on NB cells, RA in
ES cells induces neuronal differentiation, which coincides with increased expression of the α3
subunit (155). Also, CGs are thought to be more potent against cells with higher expression of
α3 to α1 subunits (155, 200). Thus if RA increases expression of α3 in NB TICs as with ES cells
then combinational treatment of RA and RIDK34 (or another CG) may exhibit synergistic
64
antitumour activity due to RA induced differentiation but also due to sensitization of NB cells to
CGs.
Conversely, RA appears to down-regulate expression of α3 in myeloid cells (HL-60), suggesting
that these and possibly other immune cells might become less susceptible to CGs in
combinatorial treatment with RA (201). However, one impediment to this treatment strategy is
that RA also induces the expression of the α3 subunit in cardiac myocytes which could
potentially increase sensitivity of cardiac myocytes and in children who may already have
decreased cardiac function this may increased the risk of cardiotoxicity. Thus this may present
one obstacle in combinatorial therapy of CGs with RA or any agent that may compromise
cardiac function. (202).
Synergism of CGs with other therapeutic agents would likely result in lower doses of CGs
required for antitumour activity thus reducing the potential for toxicity. Recently, several CGs
have been identified to have synergistic effects with drugs currently used to treat NB or are
currently under clinical trial, including oxalipatin, irinotecan, cisplatin and 5-fluoruracil. In
particular, digoxin with oxalipatin or irinotecan showed synergistic effects (159). Reports also
suggest combining selected CGs with radiotherapy as a possible therapeutic modality.
Specifically ouabain and oleandrin conferred sensitivity to radiation therapy in lung
adenocarcinoma cells (PC-3 cells) (203, 204). Furthermore, ouabain appears to selectively
increase radiosensitivity of tumour cells as it was ineffective in modifying radioresponse of
normal human lung fibroblasts (203). RIDK34 may also be a radiosensitizer such as other CGs
and may be useful in several combinational therapies. However, except with RA there is very
little literature on combination therapy of CGs in NB and may be an important future aspect to
consider when determining the utility of RIDK34 and other CGs in treating NB.
65
Chapter 6 Conclusion and Future Directions
We hypothesized that CGs are potential treatment agents for NB and by testing several CGs in
vitro including digoxin, ouabain, peruvoside and convallatoxin, we can conclude that CGs are
effective agents against NB in vitro. Also, our hypothesis was that greater antitumour activity
and lower predicted cardiotoxicity could be achieved by the modification of structural groups on
CG compounds. We proved that it is possible to increase the potency of these compounds
through different structural modifications of primarily the tetracyclic steroid ring. Our findings
demonstrate a trend at the C19 bonded group that associated drug potency against NB TICs with
polar character and steric hindrance at this site. This was demonstrated with the convallatoxin
derived series of compounds RIDK33, RIDK34 and RIDK35, as well as with the peruvoside
derived series of compounds RIDK52, RIDK53 and RIDK54. The most potent compounds in
their respective series were RIDK34 and RIDK54; both contained an oxime group bonded to the
C19 carbon. In addition to increasing polar character at this site and possibly the number of
potential hydrogen bonds, the polar character of this group is further away from the tetracyclic
steroid core than the alternative groups added to the C19 carbon, which may allow RIDK34 and
RIDK54 to have more favourable interactions with its cellular target. Structure-activity-
relationship analysis of 2″-oxovoruscharin compounds leading to the identification of
UNBS1450, showed that the addition of a hydroxyl group to the C19 carbon, in part is associated
with higher in vivo tolerance of UNBS1450 relative to its precursor compound (114). In regards
to this, RIDK34, not only displayed higher potency relative to its starting compound, but may
also be better tolerated in vivo relative to its starting compound.
Further preclinical studies are required to test the in vivo efficacy of CGs, namely RIDK34 to
treat NB. To test for in vivo efficacy, NB TICs would be injected subcutaneously into
NOD/SCID mice and after tumour formation, drug treatment would commence and tumour
response to RIDK34 would be observed. In addition, orthotropic models using NB TICs form
metastases to the spleen, adrenal gland, liver, lymph node and kidney, which would allow for
further study of the effects of CGs on migration and distant tumour formation (51). Also the
orthotropic NB model may enable studies on the effects of the microenvironment on primary
tumour and metastatic tumour response to RIDK34 and other CGs. Additionally, growing NB
66
TICs over an adherent layer of bone marrow stromal cells would allow in vitro investigation of
tumour microenvironment. Investigating the in vivo contribution to NB TIC response to CGs
would be an additional component that may inform future studies of these therapeutic agents.
Murine models could also be used to study how RIDK34 is metabolized. The parent compound
of RIDK34, convallatoxin is readily metabolized in the liver at the C19 group from an aldeyde to
an alcohol (112). Given the relative stability of the oxime group, we would suspect that RIDK34
would persist longer in the body (112); however, this would need to be confirmed by an in vivo
study. Lastly, we suspect that similar to convallatoxin, RIDK34 is readily excreted by the
kidneys and in feces. The total dose of convallatoxin is excreted in two days in healthy patients
and longer in patients with impaired renal function (112). Decreased excretion of convallatoxin
and RIDK34 may hasten CG-induced toxicity if combined with inappropriate chemotherapeutics
and will be an important consideration for future potential combinatorial drug strategies.
We lastly hypothesized that lower predicted toxicity, particularly cardiotoxicity could be
achieved by structural modification of CGs. Towards this end, Na+/K+-ATPase activity was
tested in HEK-293 cells in response to treatment with RIDK34, convallatoxin and digitoxin.
Again, we used RIDK34, our most promising compound and compared its inhibitory effect
against its parent compound convallatoxin as a surrogate measure of relative toxicity associated
with inactivation of Na+/K+-ATPase function and subsequent deregulation of [Na+]i and [K+]i.
We found that higher concentrations of RIDK34 relative to convallatoxin was required to inhibit
the Na+/K+-ATPase suggesting that it would be better tolerated in vivo and possibly display
diminished cardiotoxicity. Although informative, testing for Na+/K+-ATPase activity is not the
ideal model to assess toxic effects of RIDK34; however, given the lack of appropriate rodent
models to test this hypothesis we used in vitro Na+/K+-ATPase activity assays. In the future, to
test for adverse effects of CGs, porcine in vivo models would likely identify potential toxicity on
the gastrointestinal system, CNS and cardiovascular system. In addition to test for cardiotoxicity,
in vitro models using human cardiomyocyte progenitor cells from surgical waste that
differentiate into functional beating, and mature cardiomyocytes might be the most reliable
means to assess potential cardiotoxic effects of RIDK34 (205). These proposed toxicity models
would require testing several CGs to determine the relative concentration of RIDK34 (or other
novel CG analogues) required to induce cardiotoxicity. Lastly, this assay would be a crucial
future experiment for the development of CG drugs for further preclinical or clinical trials.
67
Evidence supports the role of the Na+/K+-ATPase as a target for CG activity, including
antitumour activity. To characterize the role of the Na+/K+-ATPase in NB cells treated with CGs,
we used three different assays. From both silencing the expression of the Na+/K+-ATPase α1
subunit and expression of a CG-insensitive Na+/K+-ATPase, our findings suggest that the
Na+/K+-ATPase plays a necessary role in CG-induced antitumour activity. However, experiments
testing the effects of increasing [K+] suggested that CGs may work through a means that does not
require the inactivation of the Na+/K+-ATPase. Therefore, we predict that CGs target the Na+/K+-
ATPase to induce antitumour activity in a manner that may not require inactivation of the
Na+/K+-ATPase. To confirm this, we would need to test if CG treatment affects the levels of
[Na+]i , [Ca2+]i and [K+]i. Additionally knockdown experiments used a pool of siRNA targeting
Na+/K+-ATPaseα1. In the future, we would likely use two siRNA sequences, separately to test
the role of the α1 subunit in CG-induced antitumour activity. Lastly, knockdown of the α1
subunit is complicated by the expression of α3 and α4 subunits in NB TICs. The sensitivity of
α3 and α4 towards CGs, as well as subunit localization, and association with signaling proteins
in NB TICs may increase the role of α3 and α4 in CG-induced antitumour activity. To assess the
contribution of these α isoforms to CG antitumour activity, selective knockdown of each may be
informative. Potentially more informative experiments in the future would also include
knockdown of the β1 subunit. This subunit associates with all four α subunits and selectively
knocking down the β1 would reduce the expression of each of these isoforms.
We observed activation of Src and ERK kinase in response to CG treatment. We speculated that
Src and ERK kinase activation does not necessarily require inactivation of Na+/K+-ATPase
activity. This was supported in the literature with reports showing that Src and ERK signal
independently of Na+/K+-ATPase inactivation (136-138, 191). To further understand the role of
Src and ERK kinase, we would need to use peptide inhibitors or drug inhibitors for these proteins
to assess whether Src and ERK signaling is necessary in NB TICs for the observed response to
digoxin and RIDK34. Alternatively, ERK1/2 may also be activated by the generation of ROS in
response to CGs. To test this theory we would need to use inhibitors of ROS (e.g. diphenylene
iodonium) followed by assessing cell death and ERK activation to elucidate the role of ROS
production in CG-induced cell death (206-208). In addition, cross talk amongst protein kinase
signaling cascades is possible and has been observed prior in several cell systems in response to
68
CG treatment (112). Thus as a starting point we would also need to look at the contribution of
JNK and p38MAPK signaling in mediating the antitumour activity of CG in NB cells.
Lastly, we observed induction of c-PARP in response to digoxin and RIDK34 treatment.
Unfortunately, we were unable to assess the process by which CG binding to the Na+/K+-ATPase
may lead to activation of c-PARP and cell death. Previous reports in the literature suggest that
CGs can induce a host of death effectors that may increase c-PARP and induce apoptosis. One
interesting target is p53, which Kulikov et al. demonstrated to be upregulated following ouabain
treatment in NB cells (128). To properly assess the role of p53 in CG-induced cell death would
require knockdown of the tumour suppressor as well as other assays that inhibit death effectors
downstream of p53. Whether potential p53 activation is downstream of Src and ERK activation
is another interesting question for further investigation. Unknown effectors of cell death are
likely to play an important role in CG-induced antitumour activity; thus, a future experiment will
be to perform a screen, possibly a phosphoproteomic screen, to identify signaling cascades that
are preferentially upregulated with CG treatment. Understanding the mechanism(s) behind CG-
induced antitumour activity will be important for future preclinical studies that may eventually
lead to clinical trials with CG derivatives to treat NB.
80 69
References 1. Marc TG, James GG, Malcolm AS, Andrew FO. Sympathetic Nervous System Tumors. 1999. p. 65-72. 2. Maris JM, Hogarty MD, Bagatell R, Cohn SL. Neuroblastoma. Lancet 2007;369(9579):2106-20. 3. Pizzo PA. Principles & Practice of Pediatric Oncology: Lippincott Williams & Wilkins. 4. De Bernardi B, Nicolas B, Boni L, et al. Disseminated neuroblastoma in children older than one year at diagnosis: comparable results with three consecutive high-dose protocols adopted by the Italian Co-Operative Group for Neuroblastoma. Journal of clinical oncology: official journal of the American Society of Clinical Oncology 2003; 21(8):1592-601. 5. Matthay KK, Villablanca JG, Seeger RC, et al. Treatment of high-risk neuroblastoma with intensive chemotherapy, radiotherapy, autologous bone marrow transplantation, and 13-cis-retinoic acid. Children's Cancer Group. The New England journal of medicine 1999;341(16):1165-73. 6. Kushner BH, Cheung NKV, LaQuaglia MP, et al. International neuroblastoma staging system stage 1 neuroblastoma: A prospective study and literature review. Journal of Clinical Oncology 1996;14(7):2174-80. 7. Stiller CA, Parkin DM. International variations in the incidence of neuroblastoma. International Journal of Cancer 1992; 52(4):538-43. 8. Knudson Jr AG, Strong LC. Mutation and cancer: neuroblastoma and pheochromocytoma. American Journal of Human Genetics 1972; 24(5):514-32. 9. Maris JM, Weiss MJ, Guo C, et al. Loss of heterozygosity at 1p36 independently predicts for disease progression but not decreased overall survival probability in neuroblastoma patients: A children's cancer group study. Journal of Clinical Oncology 2000; 18(9):1888-99. 10. Mosse YP, Laudenslager M, Khazi D, et al. Germline PHOX2B mutation in hereditary neuroblastoma [2]. American Journal of Human Genetics 2004; 75(4):727-30. 11. Trochet D, Bourdeaut F, Janoueix-Lerosey I, et al. Germline Mutations of the Paired-Like Homeobox 2B (PHOX2B) Gene in Neuroblastoma. American Journal of Human Genetics 2004; 74(4):761-4. 12. Mosse YP, Laudenslager M, Longo L, et al. Identification of ALK as a major familial neuroblastoma predisposition gene. Nature 2008; 455(7215):930-5. 13. Bourdeaut F, Trochet D, Janoueix-Lerosey I, et al. Germline mutations of the paired-like homeobox 2B (PHOX2B) gene in neuroblastoma. Cancer Lett 2005; 228(1-2):51-8. 14. Chen Y, Takita J, Choi YL, et al. Oncogenic mutations of ALK kinase in neuroblastoma. Nature 2008; 455(7215):971-4. 15. George RE, Sanda T, Hanna M, et al. Activating mutations in ALK provide a therapeutic target in neuroblastoma. Nature 2008; 455(7215):975-8. 16. Look AT, Hayes FA, Nitschke R, McWilliams NB, Green AA. Cellular DNA content as a predictor of response to chemotherapy in infants with unresectable neuroblastoma. N Engl J Med 1984; 311(4):231-5. 17. Look AT, Hayes FA, Shuster JJ, et al. Clinical relevance of tumor cell ploidy and N-myc gene amplification in childhood neuroblastoma: a Pediatric Oncology Group study. J Clin Oncol 1991; 9(4):581-91.
70
18. Brodeur GM, Seeger RC, Schwab M. Amplification of N-myc in untreated human neuroblastomas correlates with advanced disease stage. Science 1984; 224(4653):1121-4. 19. Cheung IY, Cheung NKV. Quantitation of marrow disease in neuroblastoma by real-time reverse transcription-PCR. Clinical Cancer Research 2001; 7(6):1698-705. 20. Seeger RC, Brodeur GM, Sather H. Association of multiple copies of the N-myc oncogene with rapid progression of neuroblastomas. New England Journal of Medicine 1985; 313(18):1111-6. 21. Bagatell R, Rumcheva P, London WB, et al. Outcomes of children with intermediate-risk neuroblastoma after treatment stratified by MYCN status and tumor cell ploidy. Journal of Clinical Oncology 2005; 23(34):8819-27. 22. Chan QR, Bilke S, Wei JS, et al. cDNA array-CGH profiling identifies genomic alterations specific to stage and MYCN-amplification in neuroblastoma. BMC Genomics 2004;5. 23. Stock C, Bozsaky E, Watzinger F, et al. Genes proximal and distal to MYCN are highly expressed in human neuroblastoma as visualized by comparative expressed sequence hybridization. The American journal of pathology 2008; 172(1):203-14. 24. White PS, Maris JM, Beltinger C, et al. A region of consistent deletion in neuroblastoma maps within human chromosome 1p36.2-36.3. Proceedings of the National Academy of Sciences of the United States of America 1995; 92(12):5520-4. 25. Attiyeh EF, London WB, Mossé YP, et al. Chromosome 1p and 11q deletions and outcome in neuroblastoma. New England Journal of Medicine 2005; 353(21):2243-53. 26. Maris JM, White PS, Beltinger CP, et al. Significance of chromosome 1p loss of heterozygosity in neuroblastoma. Cancer research 1995; 55(20):4664-9. 27. Bauer A, Savelyeva L, Claas A, Praml C, Berthold F, Schwab M. Smallest region of overlapping deletion in 1p36 in human neuroblastoma: a 1 Mbp cosmid and PAC contig. Genes, chromosomes & cancer 2001; 31(3):228-39. 28. Bown N, Cotterill S, Lastowska M, et al. Gain of chromosome arm 17q and adverse outcome in patients with neuroblastoma. New England Journal of Medicine 1999;340(25):1954-61. 29. Van Roy N, Vandesompele J, Berx G, et al. Localization of the 17q breakpoint of a constitutional 1;17 translocation in a patient with neuroblastoma within a 25-kb segment located between the ACCN1 and TLK2 genes and near the distal breakpoints of two microdeletions in neurofibromatosis type 1 patients. Genes Chromosomes Cancer 2002; 35(2):113-20. 30. Kaplan DR, Miller FD. Neurotrophin signal transduction in the nervous system. Current opinion in neurobiology 2000; 10(3):381-91. 31. Tanaka T, Hiyama E, Sugimoto T, Sawada T, Tanabe M, Ida N. Trk A gene expression in neuroblastoma: The clinical significance of an immunohistochemical study. Cancer 1995; 76(6):1086-95. 32. Nakagawara A, Arima-Nakagawara M, Scavarda NJ, Azar CG, Cantor AB, Brodeur GM. Association between high levels of expression of the TRK gene and favorable outcome in human neuroblastoma. The New England journal of medicine 1993; 328(12):847-54. 33. Douma S, Van Laar T, Zevenhoven J, Meuwissen R, Van Garderen E, Peeper DS. Suppression of anoikis and induction of metastasis by the neurotrophic receptor TrkB. Nature 2004; 430(7003):1034-40. 34. Nakagawara A, Azar CG, Scavarda NJ, Brodeur GM. Expression and function of TRK-B and BDNF in human neuroblastomas. Molecular and cellular biology 1994;14(1):759-67. 35. Iyer R, Evans AE, Qi X, et al. Lestaurtinib enhances the antitumor efficacy of chemotherapy in murine xenograft models of neuroblastoma. Clinical cancer research: an official journal of the American Association for Cancer Research; 16(5):1478-85.
71
36. McConville CM, Forsyth J. Neuroblastoma - A developmental perspective. Cancer letters 2003; 197(1-2):3-9. 37. Shimada H, Ambros IM, Dehner LP, et al. The International Neuroblastoma Pathology Classification (the Shimada system). Cancer 1999;86(2):364-72. 38. Sidell N. Retinoic acid-induced growth inhibition and morphologic differentiation of human neuroblastoma cells in vitro. Journal of the National Cancer Institute 1982;68(4):589-96. 39. Tsokos M, Scarpa S, Ross RA, Triche TJ. Differentiation of human neuroblastoma recapitulates neural crest development. Study of morphology, neurotransmitter enzymes, and extracellular matrix proteins. American Journal of Pathology 1987;128(3):484-96. 40. Gaetano C, Matsumoto K, Thiele CJ. In vitro activation of distinct molecular and cellular phenotypes after induction of differentiation in a human neuroblastoma cell line. Cancer research 1992;52(16):4402-7. 41. Hermann PC, Huber SL, Herrler T, et al. Distinct populations of cancer stem cells determine tumor growth and metastatic activity in human pancreatic cancer. Cell stem cell 2007;1(3):313-23. 42. Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, and cancer stem cells. Nature 2001;414(6859):105-11. 43. Ward RJ, Dirks PB. Cancer stem cells: at the headwaters of tumor development. Annual review of pathology 2007;2:175-89. 44. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nature medicine 1997;3(7):730-7. 45. Singh SK, Clarke ID, Terasaki M, et al. Identification of a cancer stem cell in human brain tumors. Cancer research 2003;63(18):5821-8. 46. Singh SK, Hawkins C, Clarke ID, et al. Identification of human brain tumour initiating cells. Nature 2004;432(7015):396-401. 47. Fang D, Nguyen TK, Leishear K, et al. A tumorigenic subpopulation with stem cell properties in melanomas. Cancer research 2005;65(20):9328-37. 48. Taylor MD, Poppleton H, Fuller C, et al. Radial glia cells are candidate stem cells of ependymoma. Cancer Cell 2005;8(4):323-35. 49. Ricci-Vitiani L, Lombardi DG, Pilozzi E, et al. Identification and expansion of human colon-cancer-initiating cells. Nature 2007;445(7123):111-5. 50. Prince ME, Sivanandan R, Kaczorowski A, et al. Identification of a subpopulation of cells with cancer stem cell properties in head and neck squamous cell carcinoma. Proceedings of the National Academy of Sciences of the United States of America 2007;104(3):973-8. 51. Hansford LM, McKee AE, Zhang L, et al. Neuroblastoma cells isolated from bone marrow metastases contain a naturally enriched tumor-initiating cell. Cancer research 2007;67(23):11234-43. 52. Mahller YY, Williams JP, Baird WH, et al. Neuroblastoma cell lines contain pluripotent tumor initiating cells that are susceptible to a targeted oncolytic virus. PLoS ONE 2009;4(1). 53. Das B, Tsuchida R, Malkin D, Koren G, Baruchel S, Yeger H. Hypoxia Enhances Tumor Stemness by Increasing the Invasive and Tumorigenic Side Population Fraction. STEM CELLS 2008;26(7):1818-30. 54. Khanna C, Jaboin JJ, Drakos E, Tsokos M, Thiele CJ. Biologically relevant orthotopic neuroblastoma xenograft models: primary adrenal tumor growth and spontaneous distant metastasis. In vivo (Athens, Greece) 2002;16(2):77-85. 55. Quintana E, Shackleton M, Sabel MS, Fullen DR, Johnson TM, Morrison SJ. Efficient tumour formation by single human melanoma cells. Nature 2008;456(7222):593-8. 56. Lee AG, ScienceDirect. ATPases. Greenwich, Conn.: JAI Press; 1996.
72
57. Kaplan JH. Biochemistry of Na,K-ATPase. 2002. p. 511-35. 58. Köksoy AA. Na+, K+-ATPase: A review. J Ankara Med Sch 2002;24(2):73-82. 59. Aizman O, Aperia A. Na,K-ATPase as a signal transducer. 2003. p. 489-96. 60. Akera T, Ng Y-C, Shieh I-S, Bero E, Brody TM, Emmett Braselton W. Effects of K+ on the interaction between cardiac glycosides and Na,K-ATPase. European journal of pharmacology 1985;111(2):147-57. 61. Askari A. Significance of protein-protein interactions to Na/K -ATPase functions+ +. Na/K-ATPase and Related ATPases 2000:17-26. 62. James PF, Grupp IL, Grupp G, et al. Identification of a Specific Role for the Na,K-ATPase α2 Isoform as a Regulator of Calcium in the Heart. Molecular cell 1999;3(5):555-63. 63. Blanco G. Na,K-ATPase subunit heterogeneity as a mechanism for tissue-specific ion regulation. Seminars in nephrology 2005;25(5):292-303. 64. Blanco G. The Na/K-ATPase and its isozymes: What we have learned using the baculovirus expression system. Frontiers in Bioscience 2005;10(SUPPL. 1):2397-411. 65. Blanco G, Mercer RW. Isozymes of the Na-K-ATPase: Heterogeneity in structure, diversity in function. American Journal of Physiology - Renal Physiology 1998;275(5 44-5). 66. Mijatovic T, Van Quaquebeke E, Delest B, Debeir O, Darro F, Kiss R. Cardiotonic steroids on the road to anti-cancer therapy. Biochimica et biophysica acta 2007;1776(1):32-57. 67. Béguin P, Crambert G, Monnet-Tschudi F, et al. FXYD7 is a brain-specific regulator of Na,K-ATPase α1-β isozymes. EMBO Journal 2002;21(13):3264-73. 68. Lubarski I, Pihakaski-Maunsbach K, Karlish SJD, Maunsbach AB, Garty H. Interaction with the Na,K-ATPase and tissue distribution of FXYD5 (related to ion channel). Journal of Biological Chemistry 2005;280(45):37717-24. 69. Skou JC. The influence of some cations on an adenosine triphosphatase from peripheral nerves. BBA - Biochimica et Biophysica Acta 1957;23(C):394-401. 70. Albers RW. Biochemical aspects of active transport. Annual Review of Biochemistry 1967;36:727-56. 71. Post RL, Hegyvary C, Kume S. Activation by adenosine triphosphate in the phosphorylation kinetics of sodium and potassium ion transport adenosine triphosphatase. The Journal of biological chemistry 1972;247(20):6530-40. 72. Blanco G. Comparison of the enzymatic properties of the Na,K-ATPase alpha 3 beta 1 and alpha 3 beta 2 isozymes. Biochemistry 1995;34(31):9897-903. 73. Blanco G, Melton RJ, Sanchez G, Mercer RW. Functional characterization of a testes-specific alpha-subunit isoform of the sodium/potassium adenosinetriphosphatase. Biochemistry 1999;38(41):13661-9. 74. Keenan SM, DeLisle RK, Welsh WJ, Paula S, Ball Jr WJ. Elucidation of the Na+, K+-ATPase digitalis binding site. Journal of Molecular Graphics and Modelling 2005;23(6):465-75. 75. Lingrel JB, Argüello JM, Van Huysse J, Kuntzweiler TA. Cation and cardiac glycoside binding sites of the Na,K-ATPase. 1997. p. 194-206. 76. Arystarkhova E, Sweadner KJ. Tissue-specific expression of the Na,K-ATPase beta 3 subunit. The presence of beta 3 in lung and liver addresses the problem of the missing subunit. Journal of Biological Chemistry 1997;272(36):22405-8. 77. Habiba A, Blanco G, Mercer RW. Expression, activity and distribution of subunits during in vitro neuronal induction. Brain research 2000;875(1-2):1-13. 78. Hieber V, Siegel GJ, Fink DJ, Beaty MW, Mata M. Differential distribution of (Na, K)-ATPase alpha isoforms in the central nervous system. Cellular and molecular neurobiology 1991;11(2):253-62.
73
79. Mata M, Siegel GJ, Hieber V, Beaty MW, Fink DJ. Differential distribution of (Na,K)-ATPase beta isoform mRNAs in the peripheral nervous system. Brain research 1991;546(1):47-54. 80. Romanovsky D, Light KE, Walker J, Dobretsov M. Target-determined expression of beta 3 isoform of the Na +,K+-ATPase in the somatic nervous system of rat. Journal of Comparative Neurology 2005;483(1):114-23. 81. Shamraj OI, Lingrel JB. A putative fourth Na+,K(+)-ATPase alpha-subunit gene is expressed in testis. Proceedings of the National Academy of Sciences of the United States of America 1994;91(26):12952-6. 82. Adams JM, Strasser A. Is tumor growth sustained by rare cancer stem cells or dominant clones? Cancer research 2008;68(11):4018-21. 83. Bagrov AY, Shapiro JI, Fedorova OV. Endogenous Cardiotonic Steroids: Physiology, Pharmacology, and Novel Therapeutic Targets. Pharmacological reviews 2009;61(1):9-38. 84. Ewart HS, Klip A. Hormonal regulation of the Na(+)-K(+)-ATPase: mechanisms underlying rapid and sustained changes in pump activity. AJP - Cell Physiology 1995;269(2):C295-311. 85. Pesce L, Guerrero C, Comellas A, Ridge KM, Sznajder JI. Beta-Agonists regulate Na,K-ATPase via novel MAPK/ERK and rapamycin-sensitive pathways. FEBS letters 2000;486(3):310-4. 86. Therien AG, Blostein R. Mechanisms of sodium pump regulation. American Journal of Physiology - Cell Physiology 2000;279(3 48-3). 87. Feraille E, Beguin P, Carranza M-L, et al. Is Phosphorylation of the alpha 1 Subunit at Ser-16 Involved in the Control of Na,K-ATPase Activity by Phorbol Ester-activated Protein Kinase C? Molecular biology of the cell 2000;11(1):39-50. 88. Lecuona E, Garcia A, Sznajder JI. A novel role for protein phosphatase 2A in the dopaminergic regulation of Na,K-ATPase. FEBS letters 2000;481(3):217-20. 89. Sznajder JI, Factor P, Ingbar DH. Invited review: Lung edema clearance: Role of Na+-K+-ATPase. Journal of applied physiology 2002;93(5):1860-6. 90. Flemming C, Lee AG, ScienceDirect. ATPases. Greenwich, Conn.: JAI Press; 1996. 91. Gonta-Grabiec K, Rossowski W, Szumiel I. Properties of Na+/K+ ATPase and alkaline phosphatase alter during spontaneous and radiation-induced leukemogenesis in mice. Neoplasma 1986;33(2):141-55. 92. Kaplan JG. Membrane cation transport and the control of proliferation of mammalian cells. Annual Review of Physiology 1978;40:19-41. 93. Shen SS, Hamamoto ST, Bern HA, Steinhardt RA. Alteration of sodium transport in mouse mammary epithelium associated with neoplastic transformation. Cancer research 1978;38(5):1356-61. 94. Weidemann H. Na/K-ATPase, endogenous digitalis like compounds and cancer development -- a hypothesis. Frontiers in bioscience : a journal and virtual library 2005;10:2165-76. 95. Munzer JS, Daly SE, Jewell-Motz EA, Lingrel JB, Blostein R. Tissue- and isoform-specific kinetic behavior of the Na,K-ATPase. Journal of Biological Chemistry 1994;269(24):16668-76. 96. Mobasheri A, Fox R, Evans I, Cullingham F, Martin-Vasallo P, Foster CS. Epithelial Na, K-ATPase expression is down-regulated in canine prostate cancer; a possible consequence of metabolic transformation in the process of prostate malignancy. Cancer cell international 2003;3(1):8. 97. Sakai H, Suzuki T, Maeda M, et al. Up-regulation of Na+,K+-ATPase Beta 3-isoform and down-regulation of the Beta 1-isoform in human colorectal cancer. FEBS letters 2004;563(1-3):151-4.
74
98. Lefranc F, Mijatovic T, Kondo Y, et al. Targeting the alpha 1 subunit of the sodium pump to combat glioblastoma cells. Neurosurgery 2008;62(1):211-21. 99. Mijatovic T, Roland I, Van Quaquebeke E, et al. The alpha 1 subunit of the sodium pump could represent a novel target to combat non-small cell lung cancers. Journal of Pathology 2007;212(2):170-9. 100. Rajasekaran AK, Rajasekaran SA. Role of Na-K-ATPase in the assembly of tight junctions. American Journal of Physiology - Renal Physiology 2003;285(3 54-3). 101. Rajasekaran SA, Ball Jr WJ, Bander NH, Liu H, Pardee JD, Rajasekaran AK. Reduced expression of beta-subunit of NA,K-ATPase in human clear-cell renal cell carcinoma. Journal of Urology 1999;162(2):574-80. 102. Rajasekaran SA, Barwe SP, Rajasekaran AK. Multiple functions of Na,K-ATPase in epithelial cells. Seminars in nephrology 2005;25(5):328-34. 103. Rajasekaran SA, Palmer LG, Quan K, et al. Na,K-ATPasebeta-subunit is required for epithelial polarization, suppression of invasion, and cell motility. Molecular biology of the cell 2001;12(2):279-95. 104. Loftås P, Onnesjö S, Widegren E, et al. Expression of FXYD-3 is an Independent Prognostic Factor in Rectal Cancer Patients With Preoperative Radiotherapy. International Journal of Radiation Oncology Biology Physics 2009;75(1):137-42. 105. Kayed H, Kleeff J, Kolb A, et al. FXYD3 is overexpressed in pancreatic ductal adenocarcinoma and influences pancreatic cancer cell growth. International Journal of Cancer 2006;118(1):43-54. 106. Geering K. FXYD proteins: New regulators of Na-K-ATPase. American Journal of Physiology - Renal Physiology 2006;290(2). 107. Geering K. Function of FXYD proteins, regulators of Na,K-ATPase. Journal of Bioenergetics and Biomembranes 2005;37(6):387-92. 108. Arimochi J, Ohashi-Kobayashi A, Maeda M. Interaction of mat-8 (FXYD-3) with Na+/K+-ATPase in colorectal cancer cells. Biological and Pharmaceutical Bulletin 2007;30(4):648-54. 109. Okudela K, Yazawa T, Ishii J, et al. Down-regulation of FXYD3 expression in human lung cancers: Its mechanism and potential role in carcinogenesis. American Journal of Pathology 2009;175(6):2646-56. 110. Haustein KO, Hentschel C, Hentschel H. The relation of membrane-ATPase activity with congestive heart failure. International Journal of Clinical Pharmacology Therapy and Toxicology 1987;25(12):651-5. 111. Scheiner-Bobis G. The sodium pump. Its molecular properties and mechanics of ion transport. European Journal of Biochemistry 2002;269(10):2424-33. 112. Newman RA, Yang P, Pawlus AD, Block KI. Cardiac glycosides as novel cancer therapeutic agents. Molecular interventions 2008;8(1):36-49. 113. Stenkvist B. Is digitalis a therapy for breast carcinoma? Oncology reports 1999;6(3):493-6. 114. Van Quaquebeke E, Simon G, Andre A, et al. Identification of a novel cardenolide (2''-oxovoruscharin) from Calotropis procera and the hemisynthesis of novel derivatives displaying potent in vitro antitumor activities and high in vivo tolerance: structure-activity relationship analyses. Journal of medicinal chemistry 2005;48(3):849-56. 115. Zhao M, Bai L, Wang L, et al. Bioactive cardenolides from the stems and twigs of Nerium oleander. Journal of natural products 2007;70(7):1098-103. 116. Ogawa H, Shinoda T, Cornelius F, Toyoshima C. Crystal structure of the sodium-potassium pump (Na+,K +-ATPase) with bound potassium and ouabain. Proceedings of the National Academy of Sciences of the United States of America 2009;106(33):13742-7.
75
117. Medarde M, Caballero E, Tomé F, et al. Synthesis and evaluation of cardiotonic activity of simple butenolides II. European journal of medicinal chemistry 1993;28(11):887-92. 118. Melero CP, Medarde M, San Feliciano A. A short review on cardiotonic steroids and their aminoguanidine analogues. Molecules 2000;5(1):51-81. 119. Paula S, Tabet MR, Ball Jr WJ. Interactions between cardiac glycosides and sodium/potassium-ATPase: Three-dimensional structure-activity relationship models for ligand binding to the E2-Pi form of the enzyme versus activity inhibition. Biochemistry 2005;44(2):498-510. 120. Katz A, Lifshitz Y, Bab-Dinitz E, et al. Selectivity of digitalis glycosides for isoforms of human Na,K-ATPase. Journal of Biological Chemistry 2010; 285(25). 121. Rahimtoola SH. Digitalis therapy for patients in clinical heart failure. Circulation 2004;109(24):2942-6. 122. Grupp G. Selective updates on mechanisms of action of positive inotropic agents. Molecular and cellular biochemistry 1987;76(2):97-112. 123. Katzung B. Basic & Clinical Pharmacology. New York: McGraw-Hill; 2007. 124. James PF, Grupp IL, Grupp G, et al. Identification of a specific role for the Na,K-ATPase alpha 2 isoform as a regulator of calcium in the heart. Molecular cell 1999;3(5):555-63. 125. Winnicka K, Bielawski K, Bielawska A, Miltyk W. Apoptosis-mediated cytotoxicity of ouabain, digoxin and proscillaridin A in the estrogen independent MDA-MB-231 breast cancer cells. Archives of Pharmacal Research 2007;30(10):1216-24. 126. Yeh JY, Huang WJ, Kan SF, Wang PS. Inhibitory effects of digitalis on the proliferation of androgen dependent and independent prostate cancer cells. Journal of Urology 2001;166(5):1937-42. 127. Kometiani P, Liu L, Askari A. Digitalis-induced signaling by Na+/K+-ATPase in human breast cancer cells. Molecular pharmacology 2005;67(3):929-36. 128. Kulikov A, Eva A, Kirch U, Boldyrev A, Scheiner-Bobis G. Ouabain activates signaling pathways associated with cell death in human neuroblastoma. Biochimica et biophysica acta 2007;1768(7):1691-702. 129. Haas M, Askari A, Xie Z. Involvement of Src and epidermal growth factor receptor in the signal-transducing function of Na+/K+-ATPase. Journal of Biological Chemistry 2000;275(36):27832-7. 130. Haas M, Wang H, Tian J, Xie Z. Src-mediated inter-receptor cross-talk between the Na+/K+-ATPase and the epidermal growth factor receptor relays the signal from ouabain to mitogen-activated protein kinases. Journal of Biological Chemistry 2002;277(21):18694-702. 131. Tian J, Cai T, Yuan Z, et al. Binding of Src to Na+/K+-ATPase forms a functional signaling complex. Molecular biology of the cell 2006;17(1):317-26. 132. Tian J, Gong X, Xie Z. Signal-transducing function of Na+-K+-atpase is essential for ouabain's effect on [Ca2+]i in rat cardiac myocytes. American Journal of Physiology - Heart and Circulatory Physiology 2001;281(5 50-5). 133. Tian J, Liu J, Garlid KD, Shapiro JI, Xie Z. Involvement of mitogen-activated protein kinases and reactive oxygen species in the inotropic action of ouabain on cardiac myocytes. A potential role for mitochondrial KATP channels. Molecular and cellular biochemistry 2003;242(1-2):181-7. 134. Wang H, Haas M, Liang M, et al. Ouabain assembles signaling cascades through the caveolar Na+/K+-ATPase. The Journal of biological chemistry 2004;279(17):17250-9. 135. Wang Z, Zheng M, Li Z, et al. Cardiac glycosides inhibit p53 synthesis by a mechanism relieved by Src or MAPK inhibition. Cancer research 2009;69(16):6556-64.
76
136. Xie Z, Askari A. Na+/K+-ATPase as a signal transducer. European Journal of Biochemistry 2002;269(10):2434-9. 137. Xie Z, Cai T. Na+-K+--ATPase-mediated signal transduction: from protein interaction to cellular function. Mol Interv 2003;3(3):157-68. 138. Liu J, Tian J, Haas M, Shapiro JI, Askari A, Xie Z. Ouabain interaction with cardiac Na+/K+-ATPase initiates signal cascades independent of changes in intracellular Na+ and Ca2+ concentrations. Journal of Biological Chemistry 2000;275(36):27838-44. 139. Karpova L, Eva A, Kirch U, Boldyrev A, Scheiner-Bobis G. Sodium pump alpha 1 and alpha 3 subunit isoforms mediate distinct responses to ouabain and are both essential for survival of human neuroblastoma. FEBS Journal;277(8):1853-60. 140. Prassas I, Paliouras M, Diamandis EP. High-throughput screening for modulators of tissue-kallikrein expression 2007. 141. Xu JW, Jin RM, Li EQ, Wang YR, Bai Y. Signal pathways in ouabain-induced proliferation of leukemia cells. World Journal of Pediatrics 2009;5(2):140-5. 142. Miyakawa-Naito A, Uhlén P P, Lal M, et al. Cell Signaling Microdomain with Na,K-ATPase and Inositol 1,4,5-Trisphosphate Receptor Generates Calcium Oscillations. Journal of Biological Chemistry 2003;278(50):50355-61. 143. Li J, Zelenin S, Aperia A, Aizman O. Low doses of ouabain protect from serum deprivation-triggered apoptosis and stimulate kidney cell proliferation via activation of NF-κB. Journal of the American Society of Nephrology 2006;17(7):1848-57. 144. Yudowski GA, Efendiev R, Pedemonte CH, Katz AI, Berggren PO, Bertorello AM. Phosphoinositide-3 kinase binds to a proline-rich motif in the Na+,K+-ATPase alpha subunit and regulates its trafficking. Proceedings of the National Academy of Sciences of the United States of America 2000;97(12):6556-61. 145. Eva A, Kirch U, Scheiner-Bobis G. Signaling pathways involving the sodium pump stimulate NO production in endothelial cells. Biochimica et Biophysica Acta- Biomembranes 2006;1758(11):1809-14. 146. Kometiani P, Tian J, Nabih Z, Gick G, Xie Z. Regulation of Na/K-ATPase β1-subunit gene expression by ouabain and other hypertrophic stimuli in neonatal rat cardiac myocytes. Molecular and cellular biochemistry 2000;215(1-2):65-72. 147. Winnicka K, Bielawski K, Bielawska A, Surazynski A. Antiproliferative activity of derivatives of ouabain, digoxin and proscillaridin A in human MCF-7 and MDA-MB-231 breast cancer cells. Biological & pharmaceutical bulletin 2008;31(6):1131-40. 148. Huang YT, Chueh SC, Teng CM, Guh JH. Investigation of ouabain-induced anticancer effect in human androgen-independent prostate cancer PC-3 cells. Biochemical pharmacology 2004;67(4):727-33. 149. Simpson CD, Mawji IA, Williams MA, Hurren R, Beheshti-Zavareh R, Schimmer AD. A high throughput cell-based screen identifies inhibitors of the Na+/K+ ATPase pump as sensitizers to anoikis and inhibitors of metastasis Cancer Res. 2009 Apr 1;69(7):2739-47. 150. Frese S, Frese-Schaper M, Andres AC, Miescher D, Zumkehr B, Schmid RA. Cardiac glycosides initiate Apo2L/TRAIL-induced apoptosis in non-small cell lung cancer cells by up-regulation of death receptors 4 and 5. Cancer research 2006;66(11):5867-74. 151. Mijatovic T, De Beeck AO, Van Quaquebeke E, et al. The cardenolide UNBS1450 is able to deactivate nuclear factor κB -mediated cytoprotective effects in human non-small cell lung cancer cells. Molecular Cancer Therapeutics 2006;5(2):391-9. 152. Mijatovic T, Mathieu V, Gaussin JF, et al. Cardenolide-induced lysosomal membrane permeabilization demonstrates therapeutic benefits in experimental human non-small cell lung cancers. Neoplasia 2006;8(5):402-12.
77
153. Newman RA, Yang P, Hittelman WN, et al. Oleandrin-mediated oxidative stress in human melanoma cells. Journal of experimental therapeutics & oncology 2006;5(3):167-81. 154. Newman RA, Kondo Y, Yokoyama T, et al. Autophagic cell death of human pancreatic tumor cells mediated by oleandrin, a lipid-soluble cardiac glycoside. Integrative cancer therapies 2007;6(4):354-64. 155. Yang P, Menter DG, Cartwright C, et al. Oleandrin-mediated inhibition of human tumor cell proliferation: importance of Na,K-ATPase alpha subunits as drug targets. Molecular cancer therapeutics 2009;8(8):2319-28. 156. Jing Y, Ohizumi H, Kawazoe N, et al. Selective inhibitory effect of bufalin on growth of human tumor cells in vitro: Association with the induction of apoptosis in leukemia HL-60 cells. Japanese Journal of Cancer Research 1994;85(6):645-51. 157. Watabe M, Masuda Y, Nakajo S, Yoshida T, Kuroiwa Y, Nakaya K. The cooperative interaction of two different signaling pathways in response to bufalin induces apoptosis in human leukemia U937 cells. Journal of Biological Chemistry 1996;271(24):14067-73. 158. Johansson S, Lindholm P, Gullbo J, Larsson R, Bohlin L, Claeson P. Cytotoxicity of digitoxin and related cardiac glycosides in human tumor cells. Anti-Cancer Drugs 2001;12(5):475-83. 159. Felth J, Rickardson L, Rosén J, et al. Cytotoxic effects of cardiac glycosides in colon cancer cells, alone and in combination with standard chemotherapeutic drugs. Journal of natural products 2009;72(11):1969-74. 160. Riganti C, Campia I, Polimeni M, Pescarmona G, Ghigo D, Bosia A. Digoxin and ouabain induce P-glycoprotein by activating calmodulin kinase II and hypoxia-inducible factor-1α in human colon cancer cells. Toxicology and applied pharmacology 2009;240(3):385-92. 161. Lefranc F, Mijatovic T, Camby I, et al. The binding of the UNBS1450 cardenolide to the sodium pump in human glioblastoma (GBM) cells dramatically impairs both their migration and proliferation properties 2006. 162. Hamed AI, Plaza A, Balestrieri ML, et al. Cardenolide glycosides from Pergularia tomentosa and their proapoptotic activity in Kaposi's sarcoma cells. Journal of natural products 2006;69(9):1319-22. 163. Yin JQ, Shen JN, Su WW, et al. Bufalin induces apoptosis in human osteosarcoma U-2OS and U-2OS methotrexate300-resistant cell lines. Acta Pharmacologica Sinica 2007;28(5):712-20. 164. Antczak C, Kloepping C, Radu C, et al. Revisiting old drugs as novel agents for retinoblastoma: in vitro and in vivo antitumor activity of cardenolides. Investigative ophthalmology & visual science 2009;50(7):3065-73. 165. Svensson A, Azarbayjani F, Bäckman U, Matsumoto T, Christofferson R. Digoxin inhibits neuroblastoma tumor growth in mice. Anticancer Research 2005;25(1 A):207-12. 166. Bielawski K, Winnicka K, Bielawska A. Inhibition of DNA topoisomerases I and II, and growth inhibition of breast cancer MCF-7 cells by ouabain, digoxin and proscillaridin A. Biological and Pharmaceutical Bulletin 2006;29(7):1493-7. 167. Chen JQ, Contreras RG, Wang R, et al. Sodium/potassium ATPase (Na+, K+-ATPase) and ouabain/related cardiac glycosides: A new paradigm for development of anti- breast cancer drugs? Breast Cancer Research & Treatment 2006;96(1):1-15. 168. Numazawa S, Shinoki MA, Ito H, Yoshida T, Kuroiwa Y. Involvement of Na+,K+-ATPase inhibition in K562 cell differentiation induced by bufalin. Journal of cellular physiology 1994;160(1):113-20. 169. Takai N, Ueda T, Nishida M, Nasu K, Narahara H. Bufalin induces growth inhibition, cell cycle arrest and apoptosis in human endometrial and ovarian cancer cells. International journal of molecular medicine 2008;21(5):637-43.
78
170. Zhang L, Nakaya K, Yoshida T, Kuroiwa Y. Induction by bufalin of differentiation of human leukemia cells HL60, U937, and ML1 toward macrophage/monocyte-like cells and its potent synergistic effect on the differentiation of human leukemia cells in combination with other inducers. Cancer research 1992;52(17):4634-41. 171. Smith JA, Madden T, Vijjeswarapu M, Newman RA. Inhibition of export of fibroblast growth factor-2 (FGF-2) from the prostate cancer cell lines PC3 and DU145 by Anvirzel and its cardiac glycoside component, oleandrin. Biochemical pharmacology 2001;62(4):469-72. 172. Zavareh RB, Lau KS, Hurren R, et al. Inhibition of the sodium/potassium ATPase impairs N-glycan expression and function. Cancer research 2008;68(16):6688-97. 173. Simpson CD, Mawji IA, Anyiwe K, et al. Inhibition of the sodium potassium adenosine triphosphatase pump sensitizes cancer cells to anoikis and prevents distant tumor formation. Cancer research 2009;69(7):2739-47. 174. Stenkvist B. Cardenolides and cancer. Anti-Cancer Drugs 2001;12(7):635-8. 175. Stenkvist B, Bengtsson E, Dahlqvist B. Cardiac glycosides and breast cancer, revisited. New England Journal of Medicine 1982;306(8):484. 176. Stenkvist B, Bengtsson E, Eklund G. Evidence of a modifying influence of heart glucosides on the development of breast cancer. Analytical and Quantitative Cytology 1980;2(1):49-54. 177. Inada A, Nakanishi T, Konoshima T, et al. Anti-tumor promoting activities of natural products. II. Inhibitory effects of digitoxin on two-stage carcinogenesis of mouse skin tumors and mouse pulmonary tumors. Biological and Pharmaceutical Bulletin 1993;16(9):930-1. 178. Mijatovic T, De Neve N, Gailly P, et al. Nucleolus and c-Myc: potential targets of cardenolide-mediated antitumor activity. Molecular cancer therapeutics 2008;7(5):1285-96. 179. F. Kayali MAJ, D. A. Laber, D. M. Miller, J. M. Day, G. H. Kloecker. Phase II trial of second-line erlotinib and digoxin in patients with non-small cell lung cancer (NSCLC). ASCO Annual Meeting; 2009; Louisville, Louisville, KY: American Society of Clinical Oncology; 2009. 180. M. I. Khan, D. A. Laber, J. Chesney, B. Taft, D. M. Miller. A phase II trial of biochemotherapy with cisplatin, vinblastine, dacarbazine, interleukin-2, interferon, and digoxin in melanoma matients. 2007. 181. Mekhail T, Kaur H, Ganapathi R, Budd GT, Elson P, Bukowski RM. Phase 1 trial of Anvirzel in patients with refractory solid tumors. Investigational new drugs 2006;24(5):423-7. 182. Lefranc F, Kiss R. The sodium pump alpha1 subunit as a potential target to combat apoptosis-resistant glioblastomas. Neoplasia (New York, NY) 2008;10(3):198-206. 183. Langenhan JM, Peters NR, Guzei IA, Hoffmann FM, Thorson JS. Enhancing the anticancer properties of cardiac glycosides by neoglycorandomization. Proceedings of the National Academy of Sciences of the United States of America 2005;102(35):12305-10. 184. S Smith KM, Datti A, Fujitani M, Grinshtein N, Zhang L, Morozova O, Blakely KM, Rotenberg SA, Hansford LM, Miller FD, Yeger H, Irwin MS, Moffat J, Marra MA, Baruchel S, Wrana JL, Kaplan DR. Selective Targeting of Neuroblastoma Tumor-Initiating Cells by Compounds Identified in a Stem Cell-Based Small Molecule Screen. EMBO Mol Med. 2010. 185. Toma JG, Akhavan M, Fernandes KJL, et al. Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nature cell biology 2001;3(9):778-84. 186. Toma JG, McKenzie IA, Bagli D, Miller FD. Isolation and characterization of multipotent skin-derived precursors from human skin. Stem cells 2005;23(6):727-37. 187. Chen QR, Bilke S, Khan J. High-resolution cDNA microarray-based comparative genomic hybridization analysis in neuroblastoma. Cancer letters 2005;228(1-2):71-81.
79
188. Oberthuer A, Skowron M, Spitz R, et al. Characterization of a complex genomic alteration on chromosome 2p that leads to four alternatively spliced fusion transcripts in the neuroblastoma cell lines IMR-5, IMR-5/75 and IMR-32. Gene 2005;363(1-2):41-50. 189. Tweddle DA, Pearson AD, Haber M, et al. The p53 pathway and its inactivation in neuroblastoma. Cancer letters 2003;197(1-2):93-8. 190. Bopp SK, Lettieri T. Comparison of four different colorimetric and fluorometric cytotoxicity assays in a zebrafish liver cell line. BMC pharmacology 2008;8:8. 191. Mohammadi K, Kometiani P, Xie Z, Askari A. Role of Protein Kinase C in the Signal Pathways That Link Na +/K+-ATPase to ERK1/2. Journal of Biological Chemistry 2001;276(45):42050-6. 192. Bielefeld EC, Hynes S, Pryznosch D, Liu J, Coleman JKM, Henderson D. A comparison of the protective effects of systemic administration of a pro-glutathione drug and a Src-PTK inhibitor against noise-induced hearing loss. Noise and Health 2005;7(29):24-30. 193. Gonfloni S, Frischknecht F, Way M, Superti-Furga G. Leucine 255 of Src couples intramolecular interactions to inhibition of catalysis. Nature structural biology 1999;6(8):760-4. 194. Akimova Oa Fau - Hamet P, Hamet P Fau - Orlov SN, Orlov SN. [Na+]i/[K+]i -independent death of ouabain-treated renal epithelial cells is not mediated by Na+,K+ -ATPase internalization and de novo gene expression. Pflugers Arch. 2008 Jan;455(4):711-9. 195. Orlov SN, Hamet P. The death of cardiotonic steroid-treated cells: Evidence of Na i +,Ki +-independent H i +-sensitive signalling. Acta Physiologica 2006;187(1-2):231-40. 196. Orlov SN, Thorin-Trescases N, Pchejetski D, et al. Na+/K+ pump and endothelial cell survival: [Na +]i/[K+]i-independent necrosis triggered by ouabain, and protection against apoptosis mediated by elevation of [Na+]i. Pflugers Archiv European Journal of Physiology 2004;448(3):335-45. 197. Manunta P, Hamilton BP, Hamlyn JM. Structure-activity relationships for the hypertensinogenic activity of ouabain: role of the sugar and lactone ring. Hypertension 2001;37(2 Part 2):472-7. 198. Karpova LV, Bulygina ER, Boldyrev AA. Different neuronal Na(+)/K(+)-ATPase isoforms are involved in diverse signaling pathways. Cell Biochem Funct. 2010 Mar;28(2):135-41. 199. Jimenez T, Sánchez, G, Wertheimer E, Blanco G. Activity of the Na,K-ATPase α4 isoform is important for membrane potential, intracellular Ca2+, and pH to maintain motility in rat spermatozoa. Reproduction;139(5):835-45. 200. Hauck C, Potter T, Bartz M, et al. Isoform specificity of cardiac glycosides binding to human Na+,K+-ATPase α1β1, α2β1 and α3β1. European journal of pharmacology 2009;622(1-3):7-14. 201. Gilmore-Hebert M, Schneider JW, Greene AL, et al. Expression of multiple Na+,K+-adenosine triphosphatase isoform genes in human hematopoietic cells. Behavior of the novel α 3 isoform during induced maturation of HL60 cells. Journal of Clinical Investigation 1989;84(1):347-51. 202. Chin S, He H, Gick G. Selective induction of Na,K-ATPase alpha 3 subunit mRNA abundance in cardiac myocytes by retinoic acid. Journal of Molecular and Cellular Cardiology 1998;30(11):2403-10. 203. Lawrence TS. Ouabain sensitizes tumor cells but not normal cells to radiation. International Journal of Radiation Oncology Biology Physics 1988;15(4):953-8. 204. Nasu S, Milas L, Kawabe S, Raju U, Newman R. Enhancement of radiotherapy by oleandrin is a caspase-3 dependent process. Cancer letters 2002;185(2):145-51. 205. Smits AM, van Vliet P, Metz CH, Korfage T, Sluijter JP, Doevendans PA, Goumans MJ. Human cardiomyocyte progenitor cells differentiate into functional mature cardiomyocytes: an in
80
vitro model for studying human cardiac physiology and pathophysiology. Nat Protoc. 2009;4(2):232-43. 206. Choudhary S, Rathore K, Wang HCR. Differential induction of reactive oxygen species through Erk1/2 and Nox-1 by FK228 for selective apoptosis of oncogenic H-Ras-expressing human urinary bladder cancer J82 cells. Journal of Cancer Research and Clinical Oncology:1-10. 207. El-Najjar N, Chatila M, Moukadem H, Vuorela H, Ocker M, Gandesiri M, Schneider-Stock R, Gali-Muhtasib H. Reactive oxygen species mediate thymoquinone-induced apoptosis and activate ERK and JNK signaling/ Apoptosis. 2010 Feb;15(2):183-95. 208. Li Z, Shi K, Guan L, et al. ROS leads to MnSOD upregulation through ERK2 translocation and p53 activation in selenite-induced apoptosis of NB4 cells. FEBS Letters;584(11):2291-7. 209. Habiba A, Mercer RW. Embryonic stem cells: A model to study Na,K-ATPase isoform expression during development. Molecular genetics and metabolism 2000;71(1-2):387-90. 210. Prassas I, Paliouras M, Datti A, Diamandis EP. High-throughput screening identifies cardiac glycosides as potent inhibitors of human tissue kallikrein expression: implications for cancer therapies. Clinical cancer research: an official journal of the American Association for Cancer Research 2008;14(18):5778-84. 211. McConkey DJ, Lin Y, Nutt LK, Ozel HZ, Newman RA. Cardiac glycosides stimulate Ca2+ increases and apoptosis in androgen-independent, metastatic human prostate adenocarcinoma cells. Cancer research 2000;60(14):3807-12. 212. Yeh JY, Huang WJ, Kan SF, Wang PS. Effects of bufalin and cinobufagin on the proliferation of androgen dependent and independent prostate cancer cells. Prostate 2003;54(2):112-24. 213. Watabe M, Nakajo S, Yoshida T, Kuroiwa Y, Nakaya K. Treatment of U937 cells with bufalin induces the translocation of Casein Kinase 2 and modulates the activity of topoisomerase II prior to the induction of apoptosis. Cell Growth and Differentiation 1997;8(8):871-9. 214. Mijatovic T, Mathieu V, De Nève N, Van Quaquebeke E, Darro F, Kiss R. Cardenolide-mediated nucleolar disorganisation contributes therapeutic benefits in subcutaneous and orthotopic xenograft models of human refractory prostate cancers 2006. 215. Perne A, Muellner MK, Steinrueck M, et al. Cardiac glycosides induce cell death in human cells by inhibiting general protein synthesis. PloS one 2009;4(12).
81
Appendix
Table 12. Na+/K+-ATPase expression profile in Neuroblastoma Tumour-Initiating Cells.
NB TIC mRNA ΔCt of Na+/K+-ATPase subunits
α1 α2 α3 α4 β1 β2 β3
NB12, p7 4.4 nd 13.7 15.1 5.2 15.8 4.4
NB88 R2, p6 4.9 nd 14 16.1 4.2 16.2 4.9
NB122R, p3 5 nd 11.2 11 4 12.3 5
Data was generously provided by from Michael Andang’s lab. Subunit levels are expressed as ΔCt target gene - ΔCt housekeeping gene β2 microglobulin. Abbreviation: nd, expression levels too low to determine.
Expression profile α and β subunits in NB TICs
Neuroblastoma is thought to be derived from a neural crest or sympathoadrenal precursor. This is
consistent with neuronal markers, Schwann cell markers and neuronal stem cell markers present
on NB tumour cells, suggesting a hierarchical structure analogous to that of neuronal stem cells
or sympathoadrenal precursors. In NB TICs the morphology is more consistent with an
undifferentiated cell type than a differentiated cell type. Also, the expression profile of α and β
subunits in NB TICs is consistent with the undifferentiated embryonic stem cells E3-ES. Similar
to E3-ES cells, NB TICs express high levels of α1 and β1, and express relatively low levels of
α2 (not expressed in E3-ES), α3 and β2; the one exception is β3 which is high in NB TICs
compared to E3-ES cells. Interestingly, mature neurons, NB tumour samples and the NB lines
SK-N-AS and SY-SY5Y all express robust levels of α3 (77, 128, 139, 209). Also, the expression
of the α3 subunit coincides temporally with neuronal maturation of E3-ES cells into neurons
following exposure to RA. This may suggest a possible role of the α3 subunit in determining a
neuronal cell fate and the association of the α3 subunit with neuronal maturation would agree
with the theory that NB TICs contain many features analogous to normal stem cells.
82
Figure 17. Crystal structure of Na+/K+-ATPase with bound ouabain. Superimposition of Cα traces of the amino acid chain of Na+/K+-ATPase from shark rectal gland. Cα traces are indicated by bound ouabain in yellow and unbound ouabain in cyan. The structure is viewed perpendicular to the surface of the cell membrane. Na+/K+-ATPase is fixed in a state analogous to E2-2K+-Pi, with MgF4
2- as a stable phosphate analogue. Ouabain (OBN; green and red) and K+ ions (I, II, and c; purple) are shown in space fill. The β-subunit, FXYD protein and several transmembrane helices are marked. From Ogawa et al.,2009 (116).
Figure 18. Reaction scheme for RIDK34. Convallatoxin (10 mg, 0.02 mmol) was dissolved in water (4 mL), added Hydroxylamine Hydrochloride (50 mg, 0.72 mmol ) and stirred at RT for 24 hrs. Upon completion, the reaction-mixture was neutralized with NaHCO3. Crude product was desalted on a reverse phase column and purified on a silica column to give RIDK34. ESMS, (ion) (expected, found); (M+Na+) (588.2788, 588.2844).
83
Table 13. Reported Mechanisms of CG antitumour activity.
Cancer Type Compound Tested Cancer Model Reported Mechanism of Action
Reference
Ouabain Peruvoside
MDA-MB-468 Inhibits kallikrein expression
(210)
proscillaridin A derivative
MDA-MB-231, MCF-7
Inhibits topoisomerase II
(147)
MDA-MB-435s Cell cycle arrest (127).
Breast
ouabain
T47D Sensitizes cells to anoikis
(173)
Oleandrin PC-3 Inhibition of FGF-2 & apoptosis and autophagy
(211), (154)
Ouabain PC-3 Apoptosis (148)
Ouabain, Peruvoside Dihydroouabain digitoxingenin digitoxinm strophanthidin
PPC-1 ouabain in vivo, PC-3 ouabain and peruvoside
Sensitizes cells to anoikis
(173)
Digoxin PPC-1 in vivo Inhibits distant tumour formation
(172)
Bufalin Cinobufagenin
LNCaP, DU145, and PC3
Apoptosis (212)
Prostate
UNBS1450 LNCaP, DU145, and PC3 in vitro & in vivo
Nucleolar disorganization apoptosis; cell cycle arrest
(178)
Glioblastoma UNBS1450 U373-MGin vitro & in vivo
Autophagy and inhibits cell migration
(98)
Melanoma Oleandrin BRO Induces oxidative stress (153)
digoxin Jurkat Cells Inhibits N-glycan biosynthesis
(172) Leukemia
Bufalin HL-60, U937, ML1, K567
Apoptosis and differentiation
(213)
Lung (NCLSC) UNBS1450 A549, A427; tested in vivo
Autophagy (151, 214)
84
oleandrin A549, Calu1, Sklu1, NCI-H6
Apo2L/TRAIL induced apoptosis
(150)
Ouabain SH-SY-5Y Apoptosis (128) Neuroblastoma Digoxin Neuro-2a Angiogenesis & tumour
growth In mice (165)
Kaposi Sarcoma Cardenolide Glycosides from Pergularia tomentosa
primary cells apoptosis (162)
Fibrosarcoma (human)
Digoxin HT1080 Inhibits cell migration/Invasion and alters N-glycan biosynthesis
(172)
Colorectoral adenocarcinoma bladder and thyroid carcinoma
Digoxin Colo320, WRO, 5637 (respectively)
Inhibits N-glycan biosynthesis
(172)
Bufalin HHUA, HEC-1B, SK-OV-3, OMC-3
Differentiation (169) Ovarian Cancer
Ouabain OVCAR3 Sensitizes cells to anoikis
(173)
Bufalin Apoptosis (163) osteosarcoma
Digitoxin
U2OS
Inhibits general protein synthesis
(215)
Retinoblastoma ouabain Y79, RB355 Apoptosis (164)
Cervical cancer Ouabain Hela Sensitizes cells to anoikis
(173)