Potential antitumor activity of digitoxin and user-designed analog administered to human lung cancer cells

Article abstract of DOI:10.1016/j.bbagen.2020.129683

Background: Cardiac glycosides (CGs), such as digitoxin, are traditionally used for treatment of congestive heart failure; recently they also gained attention for their anticancer properties. Previous studies showed that digitoxin and a synthetic L-sugar monosaccharide analog treatment decreases cancer cell proliferation, increases apoptosis, and pro-adhesion abilities; however, no reports are available on their potential to alter lung cancer cell cytoskeleton structure and reduce migratory ability. Herein, we investigated the anticancer effects of digitoxin and its analog, digitoxigenin-α-L-rhamnoside (D6MA), to establish whether cytoskeleton reorganization and reduced motility are drug-induced cellular outcomes. Methods: We treated non-small cell lung carcinoma cells (NSCLCs) with sub-therapeutic, therapeutic, and toxic concentrations of digitoxin and D6MA respectively, followed by both single point and real-time assays to evaluate changes in cellular gene and protein expression, adhesion, elasticity, and migration. Results: Digitoxin and D6MA induced a decrease in matrix metalloproteinases expression via altered focal adhesion signaling and a suppression of the phosphoinositide 3-kinases / protein kinase B pathway which lead to enhanced adhesion, altered elasticity, and reduced motility of NSCLCs. Global gene expression analysis identified dose-dependent changes to nuclear factor kappa-light-chain-enhancer, epithelial tumor, and microtubule dynamics signaling. Conclusions: Our study demonstrates that digitoxin and D6MA can target antitumor signaling pathways to alter NSCLC cytoskeleton and migratory ability to thus potentially reduce their tumorigenicity. Significance: Discovering signaling pathways that control cancer's cell phenotype and how such pathways are affected by CG treatment will potentially allow for active usage of synthetic CG analogs as therapeutic agents in advanced lung conditions.

Full text of DOI:10.1016/j.bbagen.2020.129683

Journal Pre-proof
Potential antitumor activity of digitoxin and user-designed analog
administered to human lung cancer cells
Reem Eldawud, Alixandra Wagner, Chenbo Dong, Neha Gupta,
Yon Rojanasakul, George O'Doherty, Todd A. Stueckle, Cerasela
Zoica Dinu
PII:
S0304-4165(20)30195-1
DOI:
https://doi.org/10.1016/j.bbagen.2020.129683
BBAGEN 129683
Reference:
To appear in:
BBA - General Subjects
Received date:
Revised date:
Accepted date:
24 March 2020
19 June 2020
9 July 2020
Please cite this article as: R. Eldawud, A. Wagner, C. Dong, et al., Potential antitumor
activity of digitoxin and user-designed analog administered to human lung cancer cells,
BBA - General Subjects (2020), https://doi.org/10.1016/j.bbagen.2020.129683
This is a PDF file of an article that has undergone enhancements after acceptance, such
as the addition of a cover page and metadata, and formatting for readability, but it is
not yet the definitive version of record. This version will undergo additional copyediting,
typesetting and review before it is published in its final form, but we are providing this
version to give early visibility of the article. Please note that, during the production
process, errors may be discovered which could affect the content, and all legal disclaimers
that apply to the journal pertain.
© 2020 Published by Elsevier.

Journal Pre-proof
Potential antitumor activity of digitoxin and user-designed
analog administered to human lung cancer cells
Reem Eldawuda¥, Alixandra Wagnera¥, Chenbo Donga, Neha Guptaa, Yon Rojanasakulb, George
O’Dohertyc, Todd A. Stueckled*, and Cerasela Zoica Dinua*
a Department of Chemical and Biomedical Engineering, West Virginia University, Morgantown,
WV, 26506, USA
b Department of Basic Pharmaceutical Sciences, West Virginia University, Morgantown, WV,
26506, USA
c Department of Chemistry and Chemical Biology, Northeastern University, Boston, MA, 02115,
USA
d Health Effects Laboratory Division, National Institute for Occupational Safety and Health,
Morgantown, WV, 26505, USA
1

Journal Pre-proof
¥Authors have contributed equally to this work.
*Corresponding authors:
Cerasela Zoica Dinu, Ph.D.
Department of Chemical and Biomedical Engineering
West Virginia University
Benjamin M. Statler College of Engineering and Mineral Resources
PO Box 6102
Morgantown, WV, 26506, USA
E-mail: cerasela-zoica.dinu@mail.wvu.edu
Tel.: +1 304 293 9338
Fax: +1 304 293 4139
Todd Stueckle, Ph.D.
Health Effects Laboratory Division
National Institute for Occupational Safety and Health
1095 Willowdale Road
Morgantown, WV, 26505, USA
E-mail: tstueckle@cdc.gov
Tel.: +1 304 285-6098
Fax: +1 304 285-5398
Authors Emails:
RE: reem.eldawud@gmail.com
AW: wagner.alixandra@gmail.com
CD: cbdong1985@gmail.com
NG: ngupta2@mix.wvu.edu
YR: yrojan@hsc.wvu.edu
GO: g.odoherty@neu.edu
TS: tstueckle@cdc.gov
CZD: cerasela-zoica.dinu@mail.wvu.edu
2

Journal Pre-proof
Abstract
Background: Cardiac glycosides (CGs), such as digitoxin, are traditionally used for treatment of
congestive heart failure; recently they also gained attention for their anticancer properties.
Previous studies showed that digitoxin and a synthetic L-sugar monosaccharide analog treatment
decreases cancer cell proliferation, increases apoptosis, and pro-adhesion abilities; however, no
reports are available on their potential to alter lung cancer cell cytoskeleton structure and reduce
migratory ability. Herein, we investigated the anticancer effects of digitoxin and its analog,
digitoxigenin-α-L-rhamnoside (D6MA), to establish whether cytoskeleton reorganization and
reduced motility are drug-induced cellular outcomes.
Methods: We treated non-small cell lung carcinoma cells (NSCLCs) with sub-therapeutic,
therapeutic, and toxic concentrations of digitoxin and D6MA respectively, followed by both
single point and real-time assays to evaluate changes in cellular gene and protein expression,
adhesion, elasticity, and migration.
Results: Digitoxin and D6MA induced a decrease in matrix metalloproteinases expression via
altered focal adhesion signaling and a suppression of the phosphoinositide 3-kinases / protein
kinase B pathway which lead to enhanced adhesion, altered elasticity, and reduced motility of
NSCLCs. Global gene expression analysis identified dose-dependent changes to nuclear factor
kappa-light-chain-enhancer, epithelial tumor, and microtubule dynamics signaling.
Conclusions: Our study demonstrates that digitoxin and D6MA can target antitumor signaling
pathways to alter NSCLC cytoskeleton and migratory ability to thus potentially reduce their
tumorigenicity.
General Significance: Discovering signaling pathways that control cancer’s cell phenotype and
how such pathways are affected by CG treatment will potentially allow for active usage of
synthetic CG analogs as therapeutic agents in advanced lung conditions.
Keywords
3

Journal Pre-proof
Cardiac glycosides, Digitoxin, Synthetic analog, Lung cancer, Chemotherapy, Anti-tumor
activity, Therapeutic alternatives
Abbreviations
AFM
BCA
CG
atomic force microscopy
bicinchoninic acid assay
cardiac glycoside
D6MA
DTX
DEG
E
digitoxigenin α-L-rhamnose monosaccharide
digitoxin
differentially expressed gene
Young’s modulus
ECIS
MFI
electric cell impedance sensing
mean fluorescent intensity
4

Journal Pre-proof
1. Introduction
Cardiac glycosides (CGs) are commonly used for the treatment of congestive heart failure
and arterial fibrillation (1),(2). Previous research has shown that CGs block Na+/K+-ATPase
pump activity (3) causing an increase in intracellular calcium thus allowing for more efficient
contractions to occur (4). Complementary research has also showed that CGs and user-
synthesized analogs possess anticancer activities, with studies demonstrating their selectivity
toward different cancer cells, from breast to lung cancer cell lines (2),(5),(6),(7, 8) Within this
context, we have previously compared digitoxin, a naturally occurring CG with a prolonged half-
life in human serum (7) and an L-sugar monosaccharide analog of digitoxin, digitoxigenin-α-L-
rhamnoside (D6MA) (7),(9) and identified that D6MA displayed a 5-fold increase in potency
relative to its digitoxin counterpart (6), with such anticancer activity occurring in the nanomolar
range thus helping to alleviate concerns over potential cardiotoxicity at administration (7),(10).
Our analyses were performed in non-small cell lung carcinoma cells (NSCLC) lines due to their
known resistance against drug therapies, high associated death rates relative to other cancers
(6),(7),(10), and an increased sensitivity to digitoxin and its analogs towards such cells when
compared to other studied cell lines (7),(11).
In investigating the anticancer mechanisms associated with digitoxin and its analogs, we
and others have attributed the drugs’ effects to the ability of such compounds to activate
downstream signaling pathways that regulate cellular proliferation (6),(7),(12), as well as the
increase in anoikis via Mcl-1 degradation (an anti-apoptotic protein) as associated with the GSK-
3β-mediated signaling and deregulation of the mitochondria apoptotic pathway (10). Published
reports showed that NSCLCs treated with digitoxin and its analogs displayed G1/S or G2 arrests
(6), inhibition of genes associated with cell proliferation (i.e., p53, cdc2, cyclin B1, survivin, and
Chk1/2) (6), and pro-adhesion effects through enhanced CDK4/ZONAB/ZO-1 signaling (13),
sensitization to tumor necrosis factor (TNF)-related apoptosis-inducing ligand (Apo2L/TRAIL)
(2), as well as induced cellular apoptosis (9). However, while such studies hint at the narrow
therapeutic window of CGs or synthetic analogs, they do not provide indication on the ability of
such compounds to alter cancer cell dynamics and induce their apoptosis.
We hypothesize that cellular transformation and anticancer mechanisms occur following
digitoxin and D6MA exposure in human NSCLCs through re-organization of the cancer cell
cytoskeleton and reduction in its migration capabilities. Such induced effects are envisioned to
5

Journal Pre-proof
lead to re-establishment of cell-cell and cell-substrate connections to account for a transition of
the cancerous cell to a more epithelial-like profile. To demonstrate our hypotheses we used NCI-
H460 cells as NSCLC models because of their known increased resistance to chemotherapy (14).
The selected models were treated with sub-therapeutic, therapeutic, and toxic concentrations of
digitoxin with equipotent doses of D6MA for anti-lung cancer therapy. The selectivity of
digitoxin and D6MA toward NCI-H460 cells over non-malignant human bronchial epithelial
cells (BEAS-2B) was verified using both single point and real-time assay via electric cell-
substrate impedance sensing (ECIS). The anticancer mechanisms of both digitoxin and D6MA
were evaluated via established in vitro cellular-based assays and quantitative analysis of the
changes in gene and protein expression, with the hypothesis verified for an exposure to sub-
therapeutic concentrations of both drugs. We envision that discovering the signaling pathways
that control the phenotype of cancerous cells and how such pathways are affected by exposure to
CGS or synthetic analogs will allow for active usage of these compounds as therapeutic agents
for advanced lung conditions.
6

Journal Pre-proof
2. Methods
2.1 Synthesis of D6MA
D6MA, α-L-rhamnose monosaccharide was synthesized from digitoxin (Sigma)
according to previously described methods (5),(6, 7, 10, 15-17). Briefly, the trisaccharide moiety
in digitoxin was cleaved by acid hydrolysis reaction to generate free aglycone moiety, i.e.
digitoxigenin. Subsequently, the monosaccharide group was synthesized by a palladium-
glycosylation (between digitoxigenin and a -L-BocO-pyranone 1 to 2), (18) Luche reduction
(NaBH4/CeCl3, 2 to 3) and Upjohn dihydroxylation (OsO4/NMO, 3 to D6MA) to generate
digitoxin-α-L-rhamnoside (D6MA; Supporting Information Scheme 2). Yields along the four-
step process ranged between 70 – 90%, with good overall yield of 58% (19, 20). Details on
D6MA synthesis can be found in Supporting Information.
2.2 Cell culture and treatment
Human non-small cancer lung cells (NCI-H460) and non-tumorigenic human lung
epithelial cells (BEAS-2B) were purchased from American Type Culture Collection (ATCC).
NCI-H460 cells were cultured in Roswell Park Memorial Institute (RPMI; Gibco) 1640 medium
supplemented with 10% fetal bovine serum (FBS, Life Technologies), 2 mM L-glutamine and
100-units/ml penicillin/streptomycin (Sigma). BEAS-2B cells were cultured in Dulbecco’s
Modified Eagle Medium (DMEM; Sigma) supplemented with 5% FBS, 2 mM L-glutamine and
100-units/ml penicillin/streptomycin. Cells were passaged regularly using 0.25% (w/v) trypsin
with 1.5 mM ethylenediaminetetraacetic acid (EDTA; Molecular Probes) and maintained in a
humidified atmosphere at 37ºC and 5% CO2. Stock concentrations of digitoxin and D6MA were
prepared in sterile dimethyl sulfoxide (DMSO) and diluted 1000x to obtain the exposure
concentrations used in the experiments (6). Digitoxin and D6MA exposures were performed in a
medium containing 1% FBS, 2 mM L-glutamine, and 100 units/ml penicillin/streptomycin. The
reduced concentration of FBS was adopted to circumvent existing concerns regarding potential
digitoxin or D6MA interactions with serum proteins normally present in the media (21).
7

Journal Pre-proof
2.3 Electric Cell Impedance Sensing (ECIS)
Real-time analysis of the cellular behavior was performed using an electric cell impedance
sensing system (ECIS 96W, Applied Biophysics). In each experiment, a 96W10idf ECIS array
(Applied Biophysics) containing 96 wells each with 10 gold electrodes was used to provide
concomitant measurements of 2000-4000 cells per well and at multiple frequencies. Prior to any
experiment, the ECIS gold electrodes were stabilized for 3 h in 200 µL RPMI or DMEM media
to create a reference line associated with free electrodes; subsequently, NCI-H460 or BEAS-2B
cells were seeded at a density of 2x105 cell/mL in a volume of 200 µL/well. Cells were allowed
to settle and grow over the media stabilized gold electrodes and form a confluent monolayer for
24 h. Subsequently, the cells were treated with 0, 10, 25, 50 and 100 nM digitoxin or 0, 1, 5, 10
and 25 nM D6MA and their cellular behavior post-exposure was monitored in real-time for an
additional 48 h. DMSO-only (< 0.1% v/v) treated cells served as negative controls. Different
dose ranges were chosen based on our previous work showing an about 5-fold difference in
potency on NCI-H460 toxicity between the two compounds. The resistance of the cells on the
electrodes (i.e., herein considering cells as insulators) along with the alpha parameter which
details the current through the ventral surfaces of the cells and electrodes (22) were recorded
every 12 min. Experiments were repeated 4 times with 2 replicates each, for a total of 8
replicates.
2.4 Cell size analysis
NCI-H460 cells were seeded overnight in a 12-well plates (Fisher) at a density of
2×105 cell/well and treated with digitoxin (10, 25 and 50 nM) or D6MA (1, 5 and 10 nM) after
they were allowed to form a confluent monolayer for 24 h. The cells were subsequently
collected, washed with Phosphate Buffered Saline (PBS) and analyzed for mean fluorescent
intensity (MFI) using a BD LSR Fortessa Flow cell analyzer (BD Biosciences). A calibration
curve was formed using beads with average diameters of 10.5, 14.7, 24.9 and 50 µm (Molecular
Probes). For each sample, at least 20000 events were collected using BD FACS Diva software
(Verity Software House) and analyzed using FlowJo software V10.0.7 (Tree Star Inc.).
Experiments were repeated 3 times, each with 3 replicates for a total of 9 replicates.
8

Journal Pre-proof
2.5 Biomechanical analysis
NCI-H460 cells were seeded overnight in 50 mm x 9 mm parallel culture Petri dishes (BD
Biosciences) at a density of 1 x 106 cells/dish and allowed to form a confluent monolayer.
Subsequently, the cells were treated with digitoxin (10, 25, and 50 nM) or D6MA (1, 5 and 10
nM) for 24 h. A MFP-3D-BIO atomic force microscope (AFM; Asylum Research, TE2000-U)
was used to evaluate biomechanical properties of the control (untreated) and treated cells. The
location of the AFM cantilever was confirmed using a 10x objective.
Cellular elasticity was evaluated using Sneddon’s modification of the Hertz model
developed for a four-sided pyramid (23-25). Force–displacement curves obtained during elastic
mapping were converted into force-indentation curves (26) based on the assumption that the
indented sample is extremely thick when compared to the indentation depth (27). The cells
elasticity (Young’s modulus, E) was corrected with the indentation of the tip,  through the
following equation:
 12 F
, where the Poisson’s ratio v for the cell was considered to be 0.5 (25) and
E 
2 tan 2
=36° was the open angle of the tip. The cantilever loading force F was obtained from
simultaneously recording deflection of the cantilever and multiplying that by the cantilever
spring constant (k = 0.09 N/m). Experiments were repeated 3 times; in each set of repetitions
random blocks were chosen for analysis.
2.6 Cellular adhesion and migration assay
To investigate the ability of NCI-H460 cells to adhere to and migrate was evaluated using
inserts containing porous membranes (8.0 µm pore size; Becton Dickinson). Briefly, the porous
membranes were hydrated in the cell culture media for 2 h at 37oC and 5% CO2, then the inserts
were placed in a 24 well plate containing 0.5 ml of RPMI media with 10% FBS. Cells were
suspended at a density of 1x105 cell/well in 0.5 ml of serum free media and seeded inside the
inserts for 24 h. Non-migrated cells inside the inserts were removed with a wet cotton swab. The
migrated cells on the underside of the inserts were stained and fixed by sequentially transferring
the inserts through 3 solutions provided in a manufacturer kit (Diff-Quik fixative reagent, Diff-
Quik solution I, and Diff-Quik solution II, Fisher) upon two water rinses in between. The cells’
nuclei were stained purple (Diff-Quik Solution II) while cells’ cytoplasms were stained pink
9

Journal Pre-proof
(Diff-Quik Solution I). The cell culture inserts were then allowed to air dry upside down,
overnight, and at room temperature. Images were obtained using a fluorescence microscope
(Leica Microsystems; 10X objective). Experiment was repeated 3 times, and at least 10 random
fields of view were inspected for each sample and in every trial, while counting all cells for each
experiment.
2.7 Western blot analysis
NCI-H460 cells were seeded overnight in 6-well plates (Fisher) at a density of
5×105 cell/well, and treated with digitoxin (10, 25, and 50 nM) or D6MA (1, 5, and 10 nM) for
24 h. Following exposure, the cells were lysed for 30 min in a lysis buffer containing 2% Triton
X-100 (Sigma), 1% sodium dodecyl sulfate (SDS), 100 mM sodium chloride (NaCl), 10 mM
tris-hydrochloric acid (HCl), complete mini cocktail protease inhibitors (all reagents are
purchased from Roche) and 1 mM EDTA and centrifuged at 4 °C and 12,500 rpm for 15 min.
The amount of proteins in the supernatant were determined using standard Bicinchoninic Acid
Assay (BCA, Thermo Scientific). Briefly, working reagent was prepared according to the
manufacturer instructions by mixing 50 parts of reagent A with 1 part of reagent B (reagents
included with the kit). Two μL of each sample was added to a 96-well plate and incubated with
200 μL of the working reagent at 37 °C for 30 min. Control calibration curves were prepared
using serial dilutions of standard bovine serum albumin (BSA). Absorbance at 562 nm was
recorded via the FLOUstar OPTIMA plate reader.
Protein in lysed samples and standard protein ladder (Fisher Scientific) were separated by
a 7.5 or 10 % SDS–polyacrylamide gel electrophoresis (PAGE) gel and transferred to
polyvinylidene fluoride (PVDF) membranes using a wet transfer method. Membranes were
blocked in 5% non-fat milk in distilled water for 2 h at room temperature, and subsequently
incubated with the primary antibody at 4 °C overnight. The membranes were washed 3 times in
Tris-buffered saline-Tween 20 (TBS-T; 25 mM Tris–HCl, and 125 mM NaCl, Sigma Chemicals,
and 1% Tween 20, Fisher Scientific) for 10 min each, incubated with horseradish peroxidase-
conjugated secondary antibody (Cell Signaling) for 1 h at room temperature, washed again (3
times for 10 min in TBS-T) and developed by chemiluminescence (Supersignal, ThermoFisher).
Antibodies against integrin α4, matrix metalloproteinase-2 (MMP-2), matrix metalloproteinase-9
(MMP-9), extracellular signal-regulated kinase (ERK 1/2), phosphatidylinositol 3-kinase (PI3K),
10

Journal Pre-proof
protein kinase B (AκT), and nuclear factor-kappa B (NF-κB) were investigated (Cell Signaling
Technology). Band quantification was performed using ImageJ software, version 10.2.
Experiments were performed in duplicates and repeated 3 times. Expressions levels in each
sample were normalized against expression of control β-actin (Sigma).
2.8 Global genome expression analysis
NCI-H460 cells were seeded overnight in 60 mm2 plates (Fisher) at a density of 1x106
cell/ plate. Subsequently, the cells were treated with digitoxin (10 and 25 nM), D6MA (1 and 5
nM), or DMSO vehicle control (<0.1%) for 12 h; experiments were performed in triplicate.
Global gene expression analysis of treated and control cells was performed using high-
throughput mRNA microarray analysis. For this, mRNA was collected from each group using
RNeasy Mini kits according to the manufacturer's protocol (Qiagen). Briefly, cells were lysed,
mixed with 1 volume of 70 % ethanol (Sigma), loaded onto spin column, centrifuged and washed
repeatedly. RNA elution was collected using 40 μl of nuclease-free water. RNA quality was
assessed using Eppendorf (BioPhotomer) set at 260/280 nm absorbance ratio. Upon collection,
samples were stored at − 80 °C and shipped to ArrayStar (Rockville, MD) for analysis.
Total RNA was quantified and assessed for quality with a NanoDrop ND-1000 and
denaturing agarose gel electrophoresis. RNA was reverse transcribed to cDNA with a
Superscript ds-cDNA kit (Invitrogen), labeled with one-color Cy3 DNA labeling (NimbleGen),
hybridized to a NimbleGen 12x135k human gene microarray with 45033 probes with an average
of 3 probes per target gene. Array scans were performed by Axon GenePix 4000B microarray
scanner (Molecular Devices). Images were imported to NimbleScan software (v2.5) for
alignment and expression analysis. Expression data were normalized via quantile normalization
and Robust Multichip Average algorithm to generate probe and gene level files for Agilent
GeneSpring GX (v12) analysis. Genes with 5 out of 15 samples greater than the lower cutoff
were chosen for analysis. Volcano plots with p < 0.05 (t-test) and 2-fold change in expression
compared to DMSO control determined differentially expressed genes (DEGs).
DEGs were uploaded to Ingenuity Pathway Analysis (Qiagen) to conduct pathway
analysis to thus evaluate the underlying signaling associated with the proposed hypothesis (25).
Only DEGs with known translated proteins were included. Briefly, Core Analysis was conducted
to identify the top ranked biofunctions, canonical pathways, and disease functions associated
with digitoxin and D6MA exposure to NSCLC cells. Z scores were used to predict activation or
11

Journal Pre-proof
inhibition of those genes and functions up- and downstream of observed signaling respectively.
Gene signaling networks were constructed and ranked based on likelihood of them not occurring
by chance at p < 0.05. Signaling pathways/networks for functions of interest were plotted to
evaluate underlying gene signaling to explain observed phenotypic behaviors.
2.9 Statistical analysis
Results are represented as mean ± standard deviation. Two-way analysis of variance
(ANOVA) and unpaired two-tailed Student’s t-test were performed using JMP 8.0 (SAS
Institute) and SigmaPlot 10.0 (Systat Software Inc.) to compare across treatment groups
(confirmed using T-test). The number of replicates for each experiment was detailed in their
respective section. For all experiments, results are considered significant when *p < 0.05.
12

Journal Pre-proof
3. Results and Discussion
The present study hypothesized that digitoxin, a cardiac glycoside (CG), and α
digitoxigenin-α-L-rhamnoside (D6MA), a synthetic analog (Supporting Information Scheme 1)
can reorganize malignant cell cytoskeleton and reduce their dynamics and migratory ability to
thus eliminate their tumorigenicity. To demonstrate our hypothesis, the cellular systems used as
models were non-small cell lung carcinoma cells (NSCLC) from epithelial tissue; the model was
proposed as pertinent based on previous reports that showed that 80-85% of lung cancers
originate in these cells (10) and that such cells are susceptible to our compounds (6),(7),(13).
Moreover, previous research has showed that these cells have increased resistance to
chemotherapy (14) thus leading to poor clinical and therapeutic outcomes (29). Complementary,
the cellular models to represent a model of normal, non-cancerous lung epithelial cells were
BEAS-2B cells (30).
3.1 Real time monitoring of cellular behavior following exposure to digitoxin and D6MA
We first assessed cellular behavior upon exposure of the model cells to sub-therapeutic,
therapeutic, and toxic concentrations of digitoxin (10, 25, 50, and 100 nM, respectively) and
their therapeutic concentrations of D6MA (1, 5, 10, and 25 nM), in real-time. The assessment
was based on electric cell impedance sensing (ECIS) which was previously shown to allow for
evaluation of cell proliferation (31), morphology (31),(32), and adherence (31),(32),(33), all
considering the insulating nature of the cellular membrane which causes measurable changes in
cellular resistance (31),(32),(33) onto gold electrodes (34),(35). ECIS usage was also aimed to
help discriminate possible selectivity of digitoxin and D6MA toward cancerous lung cells (7) as
well as to identify the therapeutic window for such compounds, all in a high-throughput manner
(36),(37),(31).
For the experiments, 24 h prior digitoxin and D6MA exposure cells were immobilized
onto the ECIS gold electrodes and allowed to attach, spread, and grow until the complete
formation of a cell monolayer (as indicated by a settlement in the resistance value;
Supplementary Figure S1). Attached cells were subsequently treated with sub-therapeutic,
therapeutic, and toxic concentrations of digitoxin (10, 25, 50, and 100 nM) and their equivalent
concentrations of D6MA (1, 5, 10, and 25 nM respectively); herein the sub-therapeutic,
therapeutic, and toxic concentrations were defined based on their effect on cellular
13

Journal Pre-proof
proliferation/viability as shown by previous studies (6),(7),(38). Specifically, therapeutic and
toxic doses (50 and 100 nM for digitoxin and 10 and 25 nM for D6MA, respectively) were
previously shown to cause a decrease in cellular proliferation and cell viability, whereas sub-
therapeutic concentrations caused for changes in cellular adhesion (13). Additionally, our
previous analyses have showed that the concentrations of D6MA to produce apoptosis or cell
cycle arrest in NCI-H460 cells were on average 4 to 10 times lower than the concentrations of
digitoxin that were to produce the same effects (6),(7).
Figures 1a and 1b show changes in NCI-H460 cells behavior upon exposure of such cells
to different concentrations of digitoxin and D6MA, respectively. Specifically, analysis showed
both time and dose dependent changes, with the resistance values of the cells treated with lower
concentrations of both compounds (10 and 25 nM for digitoxin, and 1 and 5 nM for D6MA
respectively) being steadily higher than the resistance values of the control cells (i.e., cells not
treated with the compounds). Furthermore, the resistance values for cells treated with 25 nM
digitoxin were higher than those for cells treated with 10 nM digitoxin. Contrary, the resistance
values for cells treated with 5 nM D6MA were higher than the values for 1 nM treated cells and
the controls, until around 24 and 42 h post exposure, respectively. Such trend subsequently
dropped to lower resistance values than those recorded for 1 nM treated cells and the control
cells, respectively. Moreover, the resistance values for cells treated with the higher
concentrations of both drugs (50 and 100 nM digitoxin and 10 and 25 nM D6MA, respectively)
underwent drastic changes when compared to the resistance values for the control cells or cells
treated with lower concentrations of these compounds. Specifically, after an initial increase, a
sharp drop was recorded around 5 h for the cells treated with concentrations of 100 nM digitoxin
and 25 nM D6MA, and around 8 h for cells treated with 50 nM digitoxin and 10 nM D6MA,
respectively.
Figures 1c and 1d showed for the first time the high-throughput and real-time screening
analysis of the normalized resistance values associated with the behavior of control BEAS-2B
cells upon their exposure to similar concentrations of both compounds. Results indicated that
exposure to high and low concentrations of digitoxin and D6MA did not lead to any major
changes in cellular resistance when compared to the resistance of controls (i.e., not treated cells),
with such results confirming previous studies reporting on compounds’ selectivity towards
neoplastic cells compared to normal cells (6).
14

Journal Pre-proof
The recorded drop in the resistance values is presumably associated with the compounds
ability to induce apoptosis and reduce cellular viability at the tested concentrations (39, 40),
presumably indicating activation of different signaling pathways that led to distinct cellular
behavior and fate of the treated group (41),(42). Specifically, analysis showed that the percentage
of apoptotic cells increased in a dose-dependent manner when NCI-H460 cells were treated with
digitoxin or D6MA (Supplementary Figure S2). Similarly, the number of live cells decreased in a
dose-dependent manner for NCI-H460 cells treated with digitoxin or D6MA (Supplementary
Figure S3). Furthermore, exposure to 100 nM digitoxin and 25 nM D6MA significantly reduced
cellular viability by more than 90%.
Overall, the differences in the resistance values between cells treated with both
compounds and across the different concentrations of the compounds being tested, suggested the
existence of a threshold concentration that would eventually lead to distinct cellular behavior and
fate. This was further confirmed when changes in cell-adhesion properties were investigated
through the use of the ECIS platform. Briefly, previous analysis showed that ECIS allows for
evaluation of the alpha parameter to reflect the restrictions on the AC current passing underneath
immobilized cells, with such parameter to be used to establish a direct relationship of cell
adhesion strengths profile (13). In particular, an increase in alpha indicated a decrease in the
distance between the cell and the electrode itself thus suggesting an increase in the cellular
adhesion properties (32),(13),(43).
Our analysis (Figure 2) showed that the alpha parameter followed a similar trend to the
resistance, with minor differences being observed across the corresponding concentrations of
both compounds being tested. Specifically, while exposure to the low concentrations of digitoxin
(10 and 25 nM) and D6MA (1 and 5 nM) resulted in an increase in alpha for the NCI-H460 cells,
with such increase persisting over time, exposure to the high concentration of both compounds
(i.e., for digitoxin 50 and 100 nM, and for D6MA 10 and 25 nM, respectively) resulted in an
increase only within the first hours of exposure. This was followed by a sharp drop, all in a dose-
dependent manner. Since the dose ranges of both compounds showed minimal effect on BEAS-
2B cells in Figure 1 and in our previous work (6),(7),(13), only alpha parameter values of NCI-
H460 cells were evaluated.
Cross comparison in the normalized alpha parameter of treated cells relative to the
control cells, indicated that exposure to 10 and 25 nM digitoxin and 1 and 5 nM D6MA
15

Journal Pre-proof
respectively, resulted in an approximate average increase of 10-15 and 12-20% after 24 and 48 h
respectively. In contrast, exposure to high concentrations resulted in a time, dose, and compound
dependent alpha decrease, where exposure to 50 nM digitoxin led to 21 and 37% decrease, and
exposure to 10 nM D6MA led to 43 and 60% decrease, both when compared to control cells after
24 and 48 h, respectively. Similarly, a 73 and 90%, and 80 and 95% decrease in the alpha
parameters was observed when cells were treated with 100 nM digitoxin and 25 nM D6MA after
24 and 48 h, respectively.
The increase in alpha recorded at low treatment concentrations suggests that both
digitoxin and D6MA promote cellular adhesion and thus hints at a less migratory behavior of
such cells to potentially resemble non-metastatic cell lines (44),(45). Contrary, the decrease in
alpha observed upon cellular treatment with high concentrations of compounds is correlated to
increased compound ability to induce apoptosis and lower cellular viability (6) thus supporting a
loss of cellular adhesion and overall integrity of the cell monolayer (46).
3.2 Effects of digitoxin and D6MA on cellular migration
Considering the above ECIS analyses showing changes in cell adhesion, we further
investigated whether such changes are influencing the ability of the treated cells to migrate. For
this, the NCI-H460 cells were treated with either of the compounds for 24 h, counted and
subsequently such live cells were used in equal amounts and immobilized onto porous
membranes for cell migration assay (28)(28)(28)(28)(28)(28)(27)(26)(25). Previous analyses
have showed that cell deformation (47),(48) and loss of focal adhesion lead to increased cell
motility (48),(49), with Gamal et al., showing that higher migration rates were associated with
cells with relatively lower adhesion values (50).
Figures 3a and 3b display optical images and quantification of the NCI-H460 migration
abilities upon treatment with both compounds. Overall, both compounds showed a decrease in
migration presumably due to induced changes in cytoskeleton or cytoskeleton-induced cellular
signaling. This hypothesis is supported by previous studies that showed that the cytoskeleton
components can act as both receptors and targets in different cellular pathways (51),(52). While
there were no differences observed between the corresponding concentrations of the compounds,
there were however significant differences being observed between treatment with therapeutic
and sub-therapeutic concentrations of both compounds. Specifically, exposure to 10, 25, and 50
16

Journal Pre-proof
nM digitoxin resulted in 25, 40, and 63 % significant reductions in the number of migrated cells
relative to migration of control cells. Similarly, cells treated with 1, 5 and 10 nM D6MA resulted
in 13, 44, and 79 % significant reduction in the number of migrated cells, again, relative to
control cells.
3.3 Effects of digitoxin and D6MA exposure on cellular size and elasticity
Changes in cytoskeleton were confirmed using cell size analysis; specifically, the volume
of the cells after 24 h of exposure to both digitoxin (10, 25, and 50 nM) and D6MA (1, 5, and 10
nM) was investigated using fluorescent assisted cell sorting (FACS) and comparing the side and
forward scatter of the suspended cells to calibration beads used as controls (36). The use of
calibration beads and forwards and side scatter plots respectively also provided a viable way to
detect the phenotype of cells and their abundance (53), thus allowing for overall information on
their physiological states (54) and height evaluations (55). For these analyses we did not
consider exposure to 100 nM digitoxin and 25 nM D6MA, respectively, due to the large
reductions observed in cellular viability and summarized in Supplementary Figure S2 and S3,
respectively.
Figure 4a shows that the average volume for all cells treated with different concentrations
of digitoxin and D6MA significantly increased by approximately 5-7% when compared to
control cells. No significant differences were however observed among the different doses or the
compounds themselves.
Secondly, cellular elasticity was investigated for confirming changes in treated cells’
cytoskeleton. Previous reports showed that changes in ECIS recorded resistance as resulted from
the changes in cell morphology, size, adhesion, and migration are likely due to cytoskeletal
alterations (31),(56) which can lead to changes in cellular elasticity (57),(58),(59) and rigidity
(47), respectively. Our analysis using atomic force microscopy (AFM) tapping mode following
24 h exposure of NCI-H460 cells to both compounds showed that the average cellular Young’s
modulus dropped by 20% for cells treated with 10 nM digitoxin, 55% for cells treated with 25
nM digitoxin, and 63% for cells treated with 50 nM digitoxin, all relative to the control cells
(Figure 4B). Contrary, the cellular elasticity for cells treated with 1 and 5 nM D6MA showed a
22 and 41% increase in Young’s elastic modulus relative to control cells, which was significantly
higher than the elasticity of the cells treated with digitoxin at the corresponding doses. Cells
17

Journal Pre-proof
treated with 10 nM D6MA showed a 90% decrease in the cellular elasticity when compared to
control cells, which was also significantly lower relative to the D6MA’s other two lower doses
being tested.
The significant alterations in Young’s modulus confirmed alterations to the cellular
cytoskeleton (57),(58) to support our above reported differences in resistance, adhesion, and cell
volume. Further, the lower values of Young’s modulus displayed by NCI-H460 cells treated with
digitoxin and 10 nM D6MA, relative to the control cells respectively, might indicate that an
individual cell gains the ability to deform (57) and potentially increase its motility (60). Cells
treated with 1 and 5 nM D6MA showed a more rigid behavior relative to control cells,
presumably reflecting their decreased ability to deform (61),(57),(62), as well as favoring a
higher adhesion, as supported by our analysis of alpha parameter.
3.4 Effects of digitoxin and D6MA on the expression of cellular regulatory proteins
Considering the observed changes in cell adhesion, migration and elasticity, the
expression levels of matrix metalloproteinases (MMPs; i.e., MMP-2 and MMP-9) and integrin
α4 were also investigated. Previous analysis showed that MMPs are a family of proteinases
involved in the proteolytic degradation of the extracellular matrix (ECM) (63),(64),(65) and
known to play major roles in cellular migration, tumor growth, and metastasis (63),(64, 66).
Furthermore, previous analyses showed that MMP-2 contributes to cancer cell migration
(63),(67),(68) through its role as a key enzyme in metastasis (67),(68), while MMP-9 mostly
mediates cell migration (69),(70),(71). Additionally, previous studies revealed that both MMP-2
and MMP-9 can be secreted in active forms to degrade ECM specific macromolecules and to
facilitate cancer progression through their involvement with MMP-14 as well as through their
overall participation in tumor angiogenesis (63). Lastly, reports also showed that overexpression
of both MMP-2 and MMP-9 was associated with epithelial to mesenchymal transition (EMT)
(63) while studies on integrin α4 were shown that the proteins has an active role in increased cell
migration and reduces cell spreading by focal adhesion formation relative to other
β1 integrins for instance (72), with Kassner et al., demonstrating that integrin’s α4 tail largely
promotes cellular migration (73).
Our analysis showed that both digitoxin and D6MA caused significant decreases in the
expression of the MMPs (Figure 5 and Supplementary Figure S4) at 24 h post treatment.
Specifically, the expression of MMP-2 decreased in a dose dependent manner upon treatment
18

Journal Pre-proof
with digitoxin, with exposure to 25 and 50 nM causing 28 and 43 % decrease relative to control
cells. Exposure to 5 and 10 nM D6MA also caused significant decreases in the expression of
MMP-2 by 54 and 50 %, respectively, again, relative to control cells. All doses of digitoxin and
D6MA showed decreases in the expression of MMP-9 relative to the control cells; however only
10 and 50 nM digitoxin and 1 and 5 nM D6MA respectively, caused significant decreases
relative to controls. Furthermore, treatment with 10, 25, and 50 nM digitoxin resulted in 42, 32,
and 54 % reductions while exposures to 1, 5, and 10 nM D6MA resulted in 32, 23, and 38 %
reductions in MMP-9 expression. There were no significant differences in the two MMPs
expression levels between the two compounds and their corresponding doses respectively.
Treatment with both digitoxin and D6MA also caused significant decreases in the
expression of integrin α4 (Figure 5) at 24 h post-treatment. Specifically, cells treated with 25 and
50 nM digitoxin displayed significant 53 and 49 % decreases in integrin α4 relative to the control
cells respectively. Cells treated with all three doses of D6MA displayed decreases in the
expression levels of integrin α4, but only the doses of 5 and 10 nM showed that such decreases
were significant when compared to the controls. Furthermore, D6MA also caused a dose-
dependent response with exposure to 1, 5, and 10 nM causing reductions in integrin α4
expression of 28, 41, and 56 %, all relative to the controls. Similar to the MMP expression, there
were no significant differences in the integrin α4 expression levels between the two compounds
and their corresponding doses respectively.
The decreased MMP expression further confirmed decreased migratory abilities of
treated cells, presumably indicating that integrin α4 might play a potential role in cell
stabilization. Such a hypothesis is supported by previous research that showed that the
expression of integrins correlates to the activation of MMPs (64),(74),(75),(76),(77), with the
interaction between integrins and MMPs being previously demonstrated to play a key role in
cellular invasion process (78) when considering their roles as substrates for MMPs (63, 78).
3.5 Global genome expression analysis
Genome expression signaling analysis found that both compounds significantly inhibited
cancer, cancer-associated respiratory disease, and malignant reproductive disease signaling (Z <
-2) at 12 h post-treatment. Several dose-dependent changes in NCI-H460 cells following
treatment with each compound were identified and were both compound and dose dependent.
19

Journal Pre-proof
Specifically, 10 nM digitoxin caused 209 differentially expressed genes (DEGs), with predicted
activation of cell signaling, immune cell trafficking, recruitment, and movement, with an
inhibition prediction of cell proliferation and colony formation respectively (Figure 6a).
Additional functions that were significantly influenced included cell morphology, organization,
development, and maintenance.
The higher dose of 25 nM digitoxin caused 769 DEGs with predicted inhibition of cell
development, morphology, and proliferation, thus further reinforcing activation of cell death,
muscle differentiation, and shape change of treated cells. Of note, inhibited cell development
included inhibition of hyperplasia, epithelial neoplasia, lung carcinoma, lung tumor, and several
leukemias. Cell movement, signaling, morphology, and organization were also significantly
impacted. Conversely, equipotent concentration of 1 nM D6MA to 10 nM digitoxin caused 1496
DEGs with predicted activation of cell movement, cytoskeleton assembly/organization,
maintenance, morphology, and skeletal and connective tissue development. Inhibited functions
included cell survival. Moreover, 5 nM D6MA exposure caused 1636 DEGs with predicted
activation of cell death, apoptosis, and necrosis through inhibition of epithelial cell quantity,
proliferation, membrane formation, plasma membrane projections, and cell activation. Similar to
25 nM digitoxin, colony formation, lung tumor, and adenoma development functions were also
inhibited.
To further evaluate both upstream and downstream signaling pathways associated with
the above behavior and concrete protein expression responses, ranked canonical pathway and
network analyses were screened. Based on behavioral and initial core analysis findings, our
detailed evaluation was restricted to 25 nM digitoxin and 5 nM D6MA. For 25 nM digitoxin, the
highest ranked network had MMP2, activated focal adhesion kinase, PLAU, OSM, and Ap1
complex associated with cardiac disease and organismal injury and abnormalities. The 4th highest
ranked network showed predicted inhibition of NF-κB complex as a hub gene for signaling as
associated with cell morphology, humoral response, and protein synthesis (Figure 6b). In
addition, the 5th highest ranked network showed hub genes down-regulated phosphatidylinositol
3-kinase (PI3K) family (PIK3C2B), an upstream effector on protein kinase B (AκT) with known
wortmannin sensitivity (79), as well as MMP14 associated with cell movement, connective
tissue disorder, and dermatological disease. Down-regulation of BCL2, an anti-apoptotic gene
typically over-expressed in cancer cells, lent further evidence to reduced AκT signaling.
20

Journal Pre-proof
Enhanced invasion signaling was noted in the actin cytoskeleton signaling pathway
primarily associated with over-expressed ITGA3 and ITGB4, receptors for ECM proteins and
MMP over-expression. However, a 105-fold downregulation of fibronectin (FN1) suggested that
over-expression of integrin mRNA may be a response to reduced ECM matrix synthesis and
reduced ability of NCI-H460 to adhere to the extracellular environment, thus confirming both
our alpha and elasticity analysis. Furthermore, NF-κB downregulation could inhibit MMP
expression. By 24 h post-treatment, decreased α4 expression suggested the potential for
decreased cellular ability and adherence to ECM. Over-expression of myosin genes along with
inactivation of PI3K, RAC, CFL, and F-actin polymerization suggested actin cytoskeleton
instability and cytoskeleton reorganization, similar to D6-MA (see below).
For 5 nM D6MA, several similarities in intracellular signaling were evident, including
over-expressed FOS, CDKN2A, THBS1, as well as down-regulated FN1 that were associated
with cell compromise, cell cycle, and cell signaling functions. Unique anti-cancer function
signaling was associated with hub genes over-expressed MDM2, NRG family, HNF4A and
under-expressed ALDH family, SOS1, KRAS, POT1, RIF1, in the top ranked networks
associated with p53, actin cytoskeleton, telomere, development and cell cycle functions. In
addition to these genes, examination of top-ranked canonical networks indicated several over-
expressed MMPs, ATM, several Erk/ Jnk protein kinases in several cancer pathways, and
decreased EGF and MET in bladder cancer and micropinocytosis pathways.
To investigate how changes in cell morphology, actin cytoskeleton, and focal adhesion
signaling relate to inhibition of epithelial neoplasia and tumors, lung cancer inhibition and lung
neoplasia inhibition networks respectively were also plotted for 25 nM digitoxin-treated NCI-
H460. Subsequently, genes associated with cell morphology were identified. Hub genes of these
large networks were similar and included over-expressed INFG, DICER1, CDKN2A, ITGA3,
MMP14, and down-regulated BCL2, FN1, LEP, ITGAX, C3, and F2 respectively; 34 out of 154
genes were associated with morphology, actin cytoskeleton, and adhesion (Supplementary Figure
S5a).
To simplify this network, we also filtered the gene expressions that were only associated
with inhibition of epithelial tumors. The resulting small signaling network included genes taking
key roles in the highest ranked signaling networks (Figure 6b, right). Over-expressed interferon
gamma (IFNG), DICER1, FOS, RUNX3, and CDKN2A/B (p16/p15) were present. INFG has
21

Journal Pre-proof
been previously shown to play important immuno-regulatory and anti-proliferative functions on
transformed cells (80).
Our finding agrees with a previous study reporting that interferons regulate
protein/protein interactions following digitoxin exposures in pancreatic cancer cells (81). CGs
along with increased interferon gamma may synergistically act on differentiating cancer cells to
earlier, more benign phenotypes, including moving leukemia cells to pre-macrophage/monocyte
phenotype (82). DICER1, a key enzyme in miRNA processing, RUNX3, and CDKN2A/B tumor
suppressor over-expression possess roles in slowing cancer cell proliferation, differentiates NCI-
H460 towards a more epithelial phenotype, and drastically improving patient survival. Such
finding correlates well with our previous publications showing S and G2 cell cycle arrest at
therapeutic digitoxin concentrations (6, 83, 84). Chen et al. (85) reported that (+)strebloside, a
natural monosaccharide tree extract, caused G2 cell cycle arrest and decreased NF-κB expression
via Na+/K+-ATPase in ovarian cancer cell lines, however, exhibited equivalent or worse potency
when compared to digitoxin. Downstream targets with anti-cancer activities included over-
expressed NF1, FOS, and THBS1, which are associated with Ras pathway inhibition, cell
differentiation, Na+/K+-ATP-ase-mediated pro-apoptotic regulatory ability, angiogenesis
inhibition, and pro-adhesive ability in lung cancer cells (86-88).
For 5 nM D6MA, examination of the 400 gene inhibition of lung tumor signaling
network revealed several cell’s structure and morphology-associated signaling changes
contributing to this network, including microtubule dynamics, actin cytoskeleton, inhibition of
membrane projections, and tumor cell shape change respectively. In particular, actin
cytoskeleton signaling in D6MA-treated NCI-H460 cells showed similar but more active effects
when compared to digitoxin, with inhibition of ROCK pathways, increased RAC signaling, and
decreased PAK1 and PAK3 expression resulting in increased focal adhesion, cytoskeleton
reorganization, myosin light chain subunits, and inhibited actin polymerization respectively
(Supplemental Figure S5b).
Microtubule dynamics signaling was prevalent in both the epithelial neoplasia and
inhibition of lung tumor signaling network. Gene expression was filtered for those genes that
either affected or decreased lung cancer (Figure 6c). This signaling network possessed over-
expressed hub genes JUN, HNF4A, THBS1, and SERPINE1, with under expressed hub genes
FN1 and EGF. Genes associated with an anti-tumor function included over-expressed THBS1
22

Journal Pre-proof
and AKAP12, with under-expressed EGF, CTNND1, WNT5A, and VHL. Such signaling
suggested cytoskeleton alterations associated with anti-angiogenesis, anti-metastatic, anti-
growth, and decreased WNT signaling (87, 89-91).
3.6 Proposed molecular pathway to support cell transformation through re-organization of the
cell cytoskeleton and reduction in its migration capabilities
Considering all the above results as well as previously published literature showing that
different pathways associated with the mediation of MMPs’ production via integrins are cell type
dependent(92), we hypothesize that digitoxin and D6MA induce cytoskeleton reorganization and
reduced motility via the PI3K/AκT pathway. These are further supported by previous studies that
showed that PI3K/AκT pathway is implicated in cell migration and invasion as well as regulation
of the transcriptional activity of NF-κB (93),(94) which has binding sites on both MMP-2 and
MMP-9 (94),(95).
Indeed, our analysis support the hypothesis and showed that all three proteins associated
with this pathway (namely PI3K, AκT, and NF-κB) were significantly affected upon cellular
exposure to digitoxin and D6MA, though there were no significant differences between the
corresponding doses of the two compounds at 24 h post-treatment (Figure 7). Specifically, a
significant decrease in PI3K was observed upon exposure of NCI-H460 cells to 10 and 50 nM
digitoxin, with decreases in expression to 37 and 71%, respectively. Further, all the doses of
D6MA caused significant decreases in PI3K expression in a dose dependent manner, with 1, 5,
and 10 nM causing decreases of 49, 64, and 75%, respectively. Cardiac glycosides are known to
bind to the plasma membrane-bound Na+/K+-ATPase to enact dose-dependent downstream
signaling pathways. Exposure at or below therapeutic concentrations does not block Na+/K+-
ATPase function, rather, the enzyme acts as a ‘signalosome’ to affect downstream pathways,
including the PI3K-mediated inhibition of Aκt phosphorylation and decreased NF-κB activation
(7, 12, 96). Our findings suggest digitoxin and D6MA bind to and activate the Na+/K+-ATPase
signalosome to affect both PI3K/Aκt and NF-κB signaling pathways.
The expression of AκT also decreased upon exposure to digitoxin and D6MA, but only at
the highest dose for each compound, with 50 nM digitoxin causing for instance a reduction of 37
% and 10 nM D6MA causing a reduction of 35 % respectively. The decrease in AκT expression
23

Journal Pre-proof
appeared to occur in a dose dependent manner for both compounds. Finally, upon exposure to
digitoxin, the expression of NF-κB decreased in a dose dependent manner with 0, 25, and 50 %
reductions for 10, 25, and 50 nM digitoxin, respectively. However, only the 50 nM caused a
significant effect relative to controls. A similar response was obtained for cells treated with
D6MA, with decreases in expression of 7, 28, and 45 % for the doses of 1, 5, and 10 nM,
respectively. NF-κB signaling is known to be influenced by cell shape, cell-to-cell contacts, and
RhoA/ROCK/myosin signaling, with lower NF-κB signaling correlating with a more epithelial-
like phenotype (97). Our results correlate to these previous findings. Furthermore, global genome
expression analysis revealed changes in cytoskeleton structure and maintenance, mainly
associated with focal adhesion, Rac, and Pak family kinase signaling pathways; the changes
clearly impacted NCI-H460 cell morphology leading to a loss in cell structure, increased
adhesion, and thus loss of mobility. Upstream changes in Rac signaling are known to impact
PAK signaling, which can influence downstream pathways of PI3K/AκT (98) currently explored
as an emerging research area in Ras-driven cancers (99, 100). Inhibition of PAK signaling can
halt carcinogenesis progression, alter cell-to-cell signaling, and reverse EMT (101). The
culmination of these structural, morphological, and adhesion-related changes were clearly
present in both digitoxin and D6MA anti-lung cancer signaling pathways presented herein.
Our conclusion of cytoskeleton reorganization and reduced migratory ability via the
PI3K/AκT pathway is also supported by previous research by Chiu et al., that showed that the
anticancer activity of H. sabdariffa leaf extract (HLE) for instance was through the AκT/NF-
κB/MMP-9 cascade pathway, with HLE causing inhibition of MMP-9 expression due to
inactivation of NF-κB as induced by a sustained inactivation of PI3K/AκT (93). Further, studies
by Wu et al., showed that the migratory abilities of hepatocellular carcinoma cells were reduced
upon exposure to sinulariolide, a cembranolide obtained from the soft coral, in a concentration-
dependent manner and again, due to suppression of the PI3K/AκT pathway and induced
inhibition of MMP-2 and-9 activities (102).
Our findings also seem to indicate that digitoxin and D6MA appear to have the same
mechanism of action, as similar trends and % reductions in viable expressed proteins were
observed for the cells treated with the two compounds. The decreased expression of integrin α4
and MMPs, along with the suppression of the PI3K/AκT pathway at 24 h post-treatment did
appear to depend on dose though, with all significant effects generally occurring at 25 and 50 nM
24

Journal Pre-proof
digitoxin and 5 and 10 nM D6MA. Additionally, the anticancer effects of the high doses for both
drugs (50 nM for digitoxin and 10 nM for D6MA) were likely occurring both via decreases in
MMPs, leading to a decrease in cellular migration as well as by induced cell apoptosis. This
effect was also observed by Wu et al., when sinulariolide was inducing hepatocarcinoma cell
apoptosis while also inhibiting metastasis (102).
Our results thus increase the existing body of evidence associated with active
mechanisms of anticancer activity for digitoxin (2),(6),(13), with such activity being attributed to
transcriptional regulatory cascades associated with the Na+/K+-ATPase signalosome (6),(7),
inhibition of genes associated with cell proliferation, cell cycle arrest (6) and increases in
adhesion due to enhanced CDK4/ZONAB/ZO-1 signaling (13), further adding the role of MMP
into reprograming the NCI-H460 cells to presumably display a more benign phenotype via
decreases in protein expression and inhibition of the resulting PI3K/AκT pathway. Our analysis
ultimately showed decreased migratory abilities and stronger adhesion characteristics, similar to
that of epithelial cells. These current findings along with current literature base were further
supported by our global gene profiling analysis that predicted inhibition of NSCLC cell
proliferation, epithelial neoplasia, and tumor growth. Collectively, these multi-target analysis and
downstream effects would drastically impact the ability of a lung tumor to proliferate, thus
suggesting the potential of digitoxin monosaccharide analog derivates as potential future
chemotherapeutic agents.
4 Conclusions
We demonstrated the antitumor ability of digitoxin and its synthetic analog, D6MA, to
cause cytoskeleton reorganization, increase adhesion, altered elasticity and reduced mobility in
human non-small cancer lung (NCI-H460) cells. Specifically, our results first showed a
selectivity of digitoxin and D6MA toward NCI-H460 cells over non-tumorigenic human lung
epithelial cells (BEAS-2B), with D6MA having around a 5x more potent effect, as confirmed
both through single point and high-throughput real-time assays. Next, exposure to the
compounds caused for increases in cell adhesion, changes in cellular volume and elasticity, and
decreases in their migratory abilities. Such effects appear to be caused by a decreased expression
of MMPs regulatory proteins, which caused the suppression of the PI3K/AκT pathway also via
decreased integrin expression. These results hint at the potential of digitoxin and D6MA to shift
25

Journal Pre-proof
the phenotype of cancerous cells toward an epithelial-like state, indicating the potential use of
CGs or synthetic analogs as therapeutic agents for the treatment of lung cancer.
Acknowledgements
The findings and conclusions in this article are those of the authors and do not necessarily
represent the view of the National Institute for Occupational Safety and Health (NIOSH),
Centers for Disease Control and Prevention. Mention of product names does not constitute
endorsement. The authors state that they have no conflicts of interest to declare.
Supporting information
This article contains supporting information submitted.
Funding
This work was supported through National Science Foundation (NSF) Grant Nos. 1434503 and
1454230, the National Institutes of Health (ES022968), and NIOSH NTRC (921043S).
Data Availability
A majority of the data that support the findings of this study are available from the corresponding
authors upon reasonable request. The global genome data that support the findings will be
available in NCBI GEO following an embargo from the date of publication.
Author contributions
R.E.D., A.W., C.D., and N.G. performed research and analyzed data; Y.R. and G.OD.-
contributed with D6MA product and technical expertise, T.A.S. and C.Z.D. designed research;
R.E.D., A.W., T.A.S. and C.Z.D. wrote the paper; all authors approved the paper.
26

Journal Pre-proof
References
1.
Osman MH, Farrag E, Selim M, Osman MS, Hasanine A, Selim A. Cardiac glycosides
use and the risk and mortality of cancer; systematic review and meta-analysis of observational
studies. PloS one. 2017;12(6):e0178611.
2.
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.
3.
the American College of Cardiology. 1985;5:16A-21A.
4. Arispe N, Diaz JC, Simakova O, Pollard HB. Heart failure drug digitoxin induces
Katz AM. Effects of Digitalis on Cell Biochemistry: Sodium Pump Inhibition. Journal of
calcium uptake into cells by forming transmembrane calcium channels. Proceedings of the
National Academy of Sciences of the United States of America. 2008;105(7):2610-5.
5.
O'Doherty G, Zhou M. The De Novo Synthesis of Oligosaccharides: Application to the
Medicinal Chemistry SAR-Study of Digitoxin. Current Topics in Medicinal Chemistry.
2008;8(2):114-25.
6.
Elbaz HA, Stueckle TA, Wang HY, O'Doherty GA, Lowry DT, Sargent LM, et al.
Digitoxin and a synthetic monosaccharide analog inhibit cell viability in lung cancer cells.
Toxicology and applied pharmacology. 2012;258(1):51-60.
7.
novel cancer therapeutics. Experimental hematology & oncology. 2012;1(1):4.
8. Ren Y, de Blanco EJC, Fuchs JR, Soejarto DD, Burdette JE, Swanson SM, et al. Potential
Anticancer Agents Characterized from Selected Tropical Plants. J Nat Prod. 2019;82(3):657-79.
9. Lo´pez-La´zaro M, Pastor N, Azrak SS, Ayuso MJ, Austin CA, Corte´s F. Digitoxin
Elbaz HA, Stueckle TA, Tse W, Rojanasakul Y, Dinu CZ. Digitoxin and its analogs as
Inhibits the Growth of Cancer Cell Lines at Concentrations Commonly Found in Cardiac
Patients. J Nat Prod 2005;68:1642-5.
10.
Pongrakhananon V, Stueckle TA, Wang HL, O'Doherty GA, Dinu CZ, Chanvorachote P,
et al. Monosaccharide digitoxin derivative sensitize human non-small cell lung cancer cells to
anoikis through Mcl-1 proteasomal degradation. Biochemical pharmacology. 2014;88(1):23-35.
11.
Iyer AK, Zhou M, Azad N, Elbaz H, Wang L, Rogalsky DK, et al. A Direct Comparison
of the Anticancer Activities of Digitoxin MeON-Neoglycosides and O-Glycosides:
Oligosaccharide Chain Length-Dependent Induction of Caspase-9-Mediated Apoptosis. ACS
medicinal chemistry letters. 2010;1(7):326-30.
12.
Therapeutic Agents. Molecular Interventions. 2008;8(1):36-49.
13. Eldawud R, Stueckle TA, Manivanna S, Elbaz H, Chen M, Rojanasakul Y, et al. Real-
Newman RA, Yang P, Pawlus AD, Block KI. Cardiac Glycosides as Novel Cancer
time analysis of the effects of toxic, therapeutic and sub-therapeutic concentrations of digitoxin
on lung cancer cells. Biosensors and Bioelectronics. 2014;59:192-9.
14.
steroids on the road to anti-cancer therapy. Biochimica et biophysica acta. 2007;1776(1):32-57.
15. Wang HY, Rojanasakul Y, O'Doherty GA. Synthesis and Evaluation of the alpha-D-
Mijatovic T, Van Quaquebeke E, Delest B, Debeir O, Darro F, Kiss R. Cardiotonic
/alpha-L-Rhamnosyl and Amicetosyl Digitoxigenin Oligomers as Anti-tumor Agents. ACS
medicinal chemistry letters. 2011;2(4):264-9.
27

Journal Pre-proof
16.
Wang H-YL, Xin W, Zhou M, Stueckle TA, Rojanasakul Y, O'Doherty GA.
Stereochemical Survey of Digitoxin Monosaccharides. ACS medicinal chemistry letters.
2010;2(1):73-8.
17.
Wang HY, Qi Z, Wu B, Kang SW, Rojanasakul Y, O'Doherty GA. C5'-Alkyl
Substitution Effects on Digitoxigenin alpha-l-Glycoside Cancer Cytotoxicity. ACS medicinal
chemistry letters. 2011;2(4):259-63.
18.
Bajaj SO, Farnsworth JR, O’Doherty GA. Enantioselective synthesis of a- and b-Boc-
protected 6-hydroxy pyranones: carbohydrate building blocks. Organic Syntheses. 2014;91:338-
55.
19.
Babu RS, O'Doherty GA. A palladium-catalyzed glycosylation reaction: the de novo
synthesis of natural and unnatural glycosides. Journal of the American Chemical Society.
2003;125(41):12406-7.
20.
medicinal chemistry SAR-study of digitoxin. Curr Top Med Chem. 2008;8(2):114-25.
21. Baggot JD, Davis LE. Plasma protein binding of digitoxin and digoxin in several
mammalian species. Res Vet Sci. 1973;15(1):81-7.
22. Lo C, Keese CR, Giaever I. Impedance Analysis of MDCK Cells Measured by Electric
Cell-Substrate Impedance Sensing. Biophysical Journal 1995;69:2800-7.
23. Rotsch C, Jacobson K, Radmacher M. Dimensional and mechanical dynamics of active
Zhou M, O'Doherty G. The de novo synthesis of oligosaccharides: application to the
and stable edges in motile fibroblasts investigated by using atomic force microscopy.
Proceedings of the National Academy of Sciences of the United States of America.
1999;96(3):921-6.
24.
Mathur AB, Truskey GA, Reichert WM. Atomic force and total internal reflection
fluorescence microscopy for the study of force transmission in endothelial cells. Biophysical
Journal. 2000;78(4):1725-35.
25.
Radmacher M, Fritz M, Kacher CM, Cleveland JP, Hansma PK. Measuring the
viscoelastic properties of human platelets with the atomic force microscope. Biophysical Journal.
1996;70(1):556-67.
26.
Laurent VM, Kasas S, Yersin A, Schaffer TE, Catsicas S, Dietler G, et al. Gradient of
rigidity in the lamellipodia of migrating cells revealed by atomic force microscopy. Biophys J.
2005;89(1):667-75.
27.
force microscopy probing of cell elasticity. Micron. 2007;38(8):824-33.
28. Wang L, Stueckle TA, Mishra A, Derk R, Meighan T, Castranova V, et al. Neoplastic-
Kuznetsova TG, Starodubtseva MN, Yegorenkov NI, Chizhik SA, Zhdanov RI. Atomic
like transformation effect of single-walled and multi-walled carbon nanotubes compared to
asbestos on human lung small airway epithelial cells. Nanotoxicology. 2014;8(5):485-507.
29.
by nanotechnology. Methods Mol Biol. 2010;596:467-88.
30. Park Y, Kim D, Dai J, Zhang Z. Human bronchial epithelial BEAS-2B cells, an
Liang XJ, Chen C, Zhao Y, Wang PC. Circumventing tumor resistance to chemotherapy
appropriate in vitro model to study heavy metals induced carcinogenesis. Toxicology and applied
pharmacology. 2015;287:240-5.
31.
technology in drug discovery. Biotechnology Journal. 2008;3:484-95.
32. Giaever I, Keese CR. A morphological biosensor for mammalian cells. Nature.
1993;366:591-2.
Xi B, Yu N, Wang X, Xu X, Abassi YA. The application of cell-based label-free
28

Journal Pre-proof
33.
Wegener J, Keese CR, Giaever I. Electric Cell-Substrate Impedance Sensing (ECIS) as a
Noninvasive Means to Monitor the Kinetics of Cell Spreading to Artificial Surfaces.
Experimental Cell Research. 2000;259(1):158-66.
34.
Sabe H. Cancer early dissemination: cancerous epithelial-mesenchymal
transdifferentiation and transforming growth factor beta signalling. Journal of biochemistry.
2011;149(6):633-9.
35.
95.
36.
Gupta GP, Massague J. Cancer metastasis: building a framework. Cell. 2006;127(4):679-
Giaever I, Keese CR. Micromotion of mammalian cells measured electrically.
Proceedings of the National Academy of Science. 1991;88:7896-900.
37. Wegener J, Keese CR, Giaever I. Electric cell-substrate impedance sensing (ECIS) as a
noninvasive means to monitor the kinetics of cell spreading to artificial surfaces. Exp Cell Res.
2000;259(1):158-66.
38.
Lopez-Lazaro M, Pastor N, Azrak SS, Ayuso MJ, Austin CA, Cortes F. Digitoxin
Inhibits the Growth of Cancer Cell Lines at Concentrations Commonly Found in Cardiac
Patients. Journal of Natural Products. 2005;68:1642-5.
39.
Eldawud R, Stueckle TA, Manivannan S, Elbaz H, Chen M, Rojanasakul Y, et al. Real-
time analysis of the effects of toxic, therapeutic and sub-therapeutic concentrations of digitoxin
on lung cancer cells. Biosensors and Bioelectronics. 2014;59:192-9.
40.
Elbaz HA, Stueckle TA, Wang H-YL, O'Doherty GA, Lowry DT, Sargent LM, et al.
Digitoxin and a synthetic monosaccharide analog inhibit cell viability in lung cancer cells.
Toxicology and applied pharmacology. 2012;258(1):51-60.
41.
Szulcek R, Bogaard HJ, van Nieuw Amerongen GP. Electric cell-substrate impedance
sensing for the quantification of endothelial proliferation, barrier function, and motility. Journal
of visualized experiments : JoVE. 2014(85).
42.
Opp D, Wafula B, Lim J, Huang E, Lo JC, Lo CM. Use of electric cell-substrate
impedance sensing to assess in vitro cytotoxicity. Biosensors & bioelectronics. 2009;24(8):2625-
9.
43.
Tiruppathi C, Malik AB, Del Vecchio PJ, Keese CR, Giaever I. Electrical method for
detection of endothelial cell shape change in real time: Assessment of endothelial barrier
function. Proceedings of the National Academy of Sciences of the United States of America.
1992;89:7919-23.
44.
May Be Indicated by Adhesion Strength. Biophys J. 2017;112(4):736-45.
45. Caron JM, Bannon M, Rosshirt L, Luis J, Monteagudo L, Sternstein GM. Methyl sulfone
Fuhrmann A, Banisadr A, Beri P, Tlsty TD, Engler AJ. Metastatic State of Cancer Cells
induces loss of metastatic properties and reemergence of normal phenotypes in a metastatic
cloudman S-91 (M3) murine melanoma cell line. PloS one. 2010;5(8):e11788.
46.
Toxicology and applied pharmacology. 2005;206(2):102-12.
47. Lautenschlager F, Paschke S, Schinkinger S, Bruel A, Beil M, Guck J. The regulatory
Xiao C, Luong JHT. Assessment of cytotoxicity by emerging impedance spectroscopy.
role of cell mechanics for migration of differentiating myeloid cells. Proceedings of the National
Academy of Sciences of the United States of America. 2009;106(37):15696-701.
48.
of adhesions. Molecular biology of the cell. 2017;28(14):1833-46.
49. Shen Y, Leng M, Yu H, Zhang Q, Luo X, Gregersen H, et al. Effect of Amphiphilic PCL-
PEG Nano-Micelles on HEpG2 Cell Migration. Macromolecular Bioscience. 2015;15:372-84.
De Pascalis C, Etienne-Manneville S. Single and collective cell migration: the mechanics
29