PreprintPDF Available
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Transformed cancer cells differ from normal cells in several important features like anchorage independence, Warburg effect and mechanosensing. Consequently, transformed cancer cells develop an anaplastic morphology and respond aberrantly to external mechanical forces. Consistent with altered mechano-responsiveness, here we show that transformed cancer cells from many different tissues have reduced growth and become apoptotic upon cyclic stretch as do normal cells after the transformation. When matrix rigidity sensing is restored in transformed cancer cells, they survive and grow faster on soft surface upon cyclic stretch like normal cells but undergo anoikis without stretch by activation of death associated protein kinase1 (DAPK1). In contrast, stretch-dependent apoptosis (mechanoptosis) of transformed cells is driven by stretch-mediated calcium influx and calcium-dependent calpain 2 protease activation on both collagen and fibronectin matrices. Further, mechanosensitive calcium channel, Piezo1 is needed for mechanoptosis. Thus, cyclic stretching of transformed cells from different tissues activates apoptosis, whereas similar stretching of normal cells stimulates growth.
Content may be subject to copyright.
1
Mechanical Stretch Kills Transformed Cancer Cells
Ajay Tijore1, Mingxi Yao1, Yu-Hsiu Wang1, Yasaman Nematbakhsh2, Anushya Hariharan1,
Chwee Teck Lim1,2,3, Michael Sheetz1,4*
1Mechanobiology Institute, National University of Singapore, Singapore 117411
2Department of Biomedical Engineering, National University of Singapore, Singapore, 117575
3Biomedical Institute for Global Health Research and Technology, National University of Singapore,
Singapore 117599
4Department of Biological Sciences, Columbia University, New York, NY10027
*Correspondence:
E-mail: ms2001@columbia.edu
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
2
Abstract
Transformed cancer cells differ from normal cells in several important features like anchorage
independence, Warburg effect and mechanosensing. Consequently, transformed cancer cells
develop an anaplastic morphology and respond aberrantly to external mechanical forces.
Consistent with altered mechano-responsiveness, here we show that transformed cancer cells
from many different tissues have reduced growth and become apoptotic upon cyclic stretch as do
normal cells after the transformation. When matrix rigidity sensing is restored in transformed
cancer cells, they survive and grow faster on soft surface upon cyclic stretch like normal cells but
undergo anoikis without stretch by activation of death associated protein kinase1 (DAPK1). In
contrast, stretch-dependent apoptosis (mechanoptosis) of transformed cells is driven by stretch-
mediated calcium influx and calcium-dependent calpain 2 protease activation on both collagen
and fibronectin matrices. Further, mechanosensitive calcium channel, Piezo1 is needed for
mechanoptosis. Thus, cyclic stretching of transformed cells from different tissues activates
apoptosis, whereas similar stretching of normal cells stimulates growth.
Keywords: malignant transformation, cell stretching, apoptosis, mechanical force, cancer
treatment, piezo calcium channel, DAPK1, calpain.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
3
Introduction
The transformed phenotype was initially described in early studies of cancer cell growth,
since tumor cells would often grow on soft agar plates, while normal cells from the same tissue
required a rigid surface for growth (Hamburger and Salmon, 1977). Recent studies have shown
that transformed cancer cells from many different tissues lack rigidity sensors. Restoration of
rigidity sensing, through cytoskeletal protein expression in cancer cells, blocks transformed
growth (Wolfenson et al., 2016; Yang et al., 2018). Further, normal cells can become
transformed by depleting cytoskeletal proteins that are required for rigidity-sensing. This raises
many questions regarding the transformed phenotype, since altering cytoskeletal protein levels in
cancer cell lines restores rigidity-dependent growth and reciprocal alterations in normal cells
cause transformed growth in many different tissue backgrounds. As the morphology and
mechanical properties of the transformed cancer cells are dramatically different from the normal
cells, there may be substantial differences in the mechano-sensitivity between normal cells and
transformed cancer cells.
Several reports have indicated that cancer cell growth is vulnerable to mechanical forces.
Studies showing inhibition of tumor growth after stretching or exercise in mice model could be
explained through a mechanical force-dependent growth inhibition (Berrueta et al., 2018; Betof
et al., 2015). Fluid shear-induced killing of circulating tumor cells and adhesive cancer cells can
also be explained by increased sensitivity to mechanical forces (Lien et al., 2013; Regmi et al.,
2017). In addition, ultrasonic and shock-wave therapies have demonstrated that cancer cells are
particularly sensitive, although concerns were raised about increased metastasis and healthy
tissue damage (Lin et al., 2012; Marano et al., 2017; Nicolai et al., 1994; Zequi et al., 2018). We
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
4
designed this study to investigate the effect of mechanical cyclic stretch on the behavior of
normal versus transformed cancer cells from a variety of different tissue backgrounds.
Because the presence or absence of a rigidity-sensing complex correlates strongly with the
normal (rigidity-dependent) and transformed growth (rigidity-independent) states, respectively, it
is relevant to understand the important elements of the rigidity sensors. They are transient
sensors that involve the assembly of sarcomeric units of ~2 µm in length with anti-parallel actin
filaments anchored to the matrix adhesion sites. Myosin IIA contraction pulls the adhesions to a
constant displacement of about 100 nm for 30 seconds irrespective of substrate rigidity
(Wolfenson et al., 2016). If the matrix is rigid, then the contractile force will exceed ~25 pN and
the adhesions will be reinforced. Adhesion disassembly and eventual cell apoptosis will occur if
the surface is soft. Many different tumor suppressors are part of the rigidity sensors including
TPM2.1 (formerly known as Tm1), myosin IIA, DAPK1, α-actinin 4 and receptor tyrosine
kinases like AXL and ROR2 (Helfman et al., 2008; Qin et al., 2018; Yang et al., 2016; Yang et
al., 2018). A common mechanism of the malignant transformation is the depletion of TPM2.1 or
the increased expression of Tpm3 (Raval et al., 2003; Stehn et al., 2013). Decreasing the ratio of
TPM2.1 to Tpm3 causes the loss of matrix rigidity sensing and enables transformed cell growth.
This process can occur even in the normal cells (Wolfenson et al., 2016; Yang et al., 2018). In
contrast, restoration of TPM2.1 level in many cancer cells re-establishes rigidity-dependent
growth and inhibits transformation in those cells (Yang et al., 2018). Other mechanosensory
cytoskeletal proteins (myosin IIA, α-actinin, filamin A, AXL, ROR) are part of the rigidity
sensing complex. Depletion of these sensory proteins and other proteins like scaffolding proteins
(caveolin1) promotes cell transformation, while their restoration in cancer cells will block
transformed cell growth (Lin et al., 2015; Wolfenson et al., 2016; Yang et al., 2016; Yang et al.,
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
5
2018). Thus, the transformed state of cancer cells appears in many different tissue backgrounds
to result from the loss of rigidity sensing, i.e. transformed cells grow on soft surfaces because
they fail to sense the surface softness.
There are a number of common features of cancer cell mechanics that differ from normal
cells, including a decrease in the rigidity of the cortical cytoskeleton (Lin et al., 2015;
Swaminathan et al., 2011) and increased traction forces on matrices (Kraning-Rush et al., 2012).
In addition, there are multiple reports demonstrating the altered expression and function of
calcium channels as well as increased calpain activity in cancer cells (Azimi et al., 2014; Storr et
al., 2011). In this study, we explore various aspects of the transformed cancer cells that cause
selective apoptosis of the transformed, but not of the normal cells. Surprisingly, we find that the
cyclic stretching of transformed cancer cells inhibits their growth and activates apoptosis.
Restoration of matrix rigidity sensing in transformed cancer cells by TPM2.1 expression results
in increased growth and survival upon cyclic stretch, particularly on soft surfaces. Transformed
cell death results from increased calcium entry upon stretch that activates calpain 2 mediated-
mitochondrial apoptosis.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
6
Results
Cancer Cells Elongate upon Cyclic Stretch: Magnitude, Frequency and Rigidity
Dependence
To understand the impact of cyclic stretch on cancer cell behavior, we used the MDA-MB-231
metastatic breast cancer cell line which is known for being highly aggressive. Cyclic stretching
of MDA-MB-231 cells on fibronectin-coated flat PDMS (2 MPa, rigid) for 6 hrs caused
prominent cell elongation (quantified by the cell aspect ratio (AR)) (Figures 1A and S2A). In the
case of MDA-MB-231 cells, the highest AR was at 5% cyclic strain (AR ~12) in contrast to 1%
(AR ~3) and non-stretched cells (AR ~2). In addition, a single stretch (5% for 6 hrs) did not alter
cell elongation (AR was the same as non-stretched cells). Next, to assess the effect of stretching
frequency on MDA-MB-231 cell morphology, cells were subjected to different frequencies (0.1
Hz to 1 Hz) for 6 hrs. Interestingly, cells displayed frequency-dependent cell elongation with
maximum elongation observed at 0.5 Hz with 5% cyclic strain (AR ~11) (Figure 1B). On the
other hand, cells showed less elongation at 0.1 Hz (AR ~4) and no significant difference in the
elongation for 1 Hz vs. 0.5 Hz was observed. Thus, the cancer cell elongation in response to
cyclic stretch optimally at a frequency of 0.5 Hz and a strain of 5% was chosen to minimize
damage to normal cells.
To investigate the effect of matrix rigidity on stretch-dependent cancer cell elongation,
MDA-MB-231 cells were cyclically stretched (5%, 0.5 Hz) on surfaces with different rigidity
(Figure 1C). On soft pillars (~8 kPa), cells failed to display elongation in response to cyclic
stretch and showed similar AR values on stretch and non stretch surfaces (Figure 1D). In fact,
cells were round on soft surfaces which resembled the morphology of non-stretched cells on soft
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
7
surfaces. In contrast on rigid pillars (~55 kPa) and flat PDMS surfaces (2 MPa), cells had an
elongated morphology, while non-stretched cells had a non-polarized morphology on both
surfaces. In addition, we observed that cyclic stretching was responsible for a small increase in
the cell area irrespective of the matrix rigidity (Figure 1D).
To delineate the influence of extracellular matrix protein (ECM) on stretch-dependent
cancer cell morphology, similar stretching experiments were repeated with MDA-MB-231 cells
cultured on collagen-I coated surfaces (Figure S3A). Again, cells elongated after cyclic stretch
on flat PDMS surfaces (AR ~9) in comparison to non-stretched cells (AR ~3). In contrast,
stretching did not result in the cell elongation on soft surfaces (AR ~3). Further, the stretch-
dependent cell area increment was noticed on rigid surfaces only. In short, MDA-MB-231 cells
responded similarly on fibronectin and collagen-I coated surfaces by showing elongated and
round morphology on rigid and soft surfaces respectively upon cyclic stretch (Figure 1E).
Rigidity-Dependent Cancer Cells Spread like Normal Cells upon Cyclic Stretch
MDA-MB-231 cells lack the cytoskeletal protein TPM2.1 which is required for proper
rigidity sensing (Wolfenson et al., 2016). When TPM2.1 was restored to MDA-MB-231 cells
(rigidity-dependent cancer cells) and they were cyclically stretched (5%, 0.5 Hz) for 6 hrs,
spread area increased with no change in AR in stark contrast to the wild-type MDA-MB-231
(Figures 2A and S2B) (AR~3 for both non-stretched and stretched MDA-MB-231-TPM2.1
cells). Further, we tested whether cancer cells from another tissue, SKOV3 (human ovarian
adenocarcinoma which lacks endogenous TPM 2.1) would behave similarly to MDA-MB-231
cells. SKOV3 cells also demonstrated cyclic stretch-dependent cell elongation (AR~7) after 6 hrs
compared to the non-stretched SKOV3 cells (AR~3) (Figure 2B). No increase in the cell area
with stretch was observed for SKOV3 cells. However, when TPM2.1 was expressed in SKOV3
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
8
cells, they spread more symmetrically (AR~3) and had a 2.5 fold increase in cell area upon
stretch when compared with wild-type SKOV3 (Figure 2C). Thus, rigidity-dependent MDA-MB-
231 and SKOV3 cells (with restored level of TPM2.1) behaved like wild-type mouse embryonic
fibroblasts (MEFs) and breast epithelial cells (MCF10A) upon cyclic stretch (Figure 2D and
S4A).
To determine if the loss of TPM 2.1, which caused transformation in MEFs and MCF10A,
would cause these cells to behave similarly to the transformed cancer cells, we knocked down
TPM2.1 in those cells (Figure 2E & S4B). Although wild-type-MEFs and -MCF10A showed a
significant increase in spread area with no increase in AR after 6 hrs of cyclic stretching (5%, 0.5
Hz), TPM2.1 KD-MEFs and MCF10As were elongated upon stretch (Figure 2D and S4A).
Statistical analysis confirmed that there was no significant difference in AR values of wild-type
cells before and after stretch, whereas TPM 2.1 KD cells demonstrated a two-fold increase in AR
after stretch (Figure 2F and S4A). Moreover, there was a stretch-dependent increase in cell area
in both wild-type and TPM2.1 KD MEF cells similar to MDA-MB-231 cells after stretch. On the
other hand, in MCF10A, we only observed a stretch-dependent cell area increment in wild-type
cells. Overall, these results indicate that the cytoskeletal protein TPM2.1 is required for proper
sensing of the external mechanical forces that promoted cell spreading in response to the periodic
stretching (Figure 2G).
Cyclic Stretch Reduces Cancer Cell Growth but Stimulates Normal Cell Growth
MDA-MB-231 cells are known for aggressive growth behavior. Therefore, we tested the
effect of cyclic stretch on MDA-MB-231, MDA-MB-231-TPM2.1 and normal cell growth
(Figure 3A). Unlike previous studies of cyclic stretching effects on fibroblast growth (Cui et al.,
2015), cyclic stretching (5%, 0.5 Hz) of MDA-MB-231 cells for 9 hours significantly reduced
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
9
their proliferation on fibronectin coated rigid and soft surfaces (Figure 3B and 3C). Similarly,
cyclic stretch-dependent inhibition of proliferation was observed on the collagen-coated surfaces
(Figure S3B). However, a dramatic shift in the effect of stretch on proliferation was found when
MDA-MB-231-TPM2.1 cells were stretched. MDA-MB-231-TPM2.1 cells exhibited a higher
proliferation rate with cyclic stretch than without on both rigid and soft surfaces. When a similar
experiment was repeated using normal cells like MEFs and MCF10A as the control cells, these
cells also had a higher proliferation rate (similar to the MDA-MB-231-TPM2.1 cells) with cyclic
stretch than without on both surfaces (Figure 3B, 3C and S4C). After MEFs were transformed by
TPM2.1 KD, cyclic stretch significantly inhibited proliferation in comparison with the non-
stretched cells. Thus, transformation of MEFs by TPM2.1 KD caused a similar behavior to
transformed MDA-MB-231 cancer cells with cyclic stretch, indicating that cells in the
transformed state had increased sensitivity to stretch-induced growth inhibition. In contrast,
normal cells and TPM2.1 expressing MDA-MB-231 cells displayed elevated growth upon cyclic
stretch.
To further verify the importance of TPM2.1 in regulating cell proliferation, we analyzed
SKOV3 cells (no endogenous TPM2.1). Interestingly, cyclic stretch again caused a reduction in
the proliferation of SKOV3 cells on rigid surfaces, while the proliferation rate increased in
TPM2.1 expressing SKOV3 cells with cyclic stretch (Figure S5). Thus, transformed cell growth
was inhibited by cyclic stretch, whereas in cancer cells that were normalized by TPM2.1
expression, cyclic stretch activated growth.
HT1080 (human fibrosarcoma) cells exhibited normal levels of TPM2.1 but were
transformed because of high levels of Tpm3 that competed with TPM2.1 and blocked rigidity
sensing (Gateva et al., 2017; Yang et al., 2018). As predicted from other transformed cells
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
10
results, cyclic stretch decreased the proliferation rate of HT1080 cells on rigid and soft surfaces
(Figure 3D). Thus for several different cell lines from different tissues, cyclic stretch inhibited
transformed cell growth and promoted normal cell growth (Figure 3E).
Cyclic Stretch Triggers Cancer Cell Apoptosis but Protects Normal Cells
To determine if inhibition of growth correlated with an increase in cell apoptosis, cells
were analyzed for apoptosis after 24 hrs of cyclic stretch by annexin-V immunostaining. As
shown in Figure 4A, cyclic stretch activated substantial apoptosis in MDA-MB-231 cells on rigid
and particularly on soft surfaces. Closer inspection of apoptotic cells confirmed that they
exhibited the characteristics of apoptosis such as membrane blebbing and cell rounding.
Interestingly, the apoptosis rate was much higher on the soft surface (34%) where cyclic stretch
did not cause cell elongation as compared to the rigid one (18%) (Figure 4B). We observed a
similar trend of apoptosis on collagen-coated rigid and soft surfaces (Figure S3C). Further, after
transformation of MEFs and MCF10A cells by TPM2.1 KD, cyclic stretching caused a dramatic
increase in apoptosis on rigid (20% for MEFs and 54% for MCF10A) and soft (35% for MEFs
and 84% for MCF10A) surfaces when compared with the non-stretched control (Figure 4A, 4B
and S4D). To strengthen the significance of cyclic stretch in promoting transformed cell
apoptosis, MDA-MB-231 cell apoptosis was compared after 24 and 48 hrs of cyclic stretch. We
found a notably higher extent of apoptosis (37-39%) after 48 hours on both rigid and soft
surfaces (Figure S6) that was likely an underestimate, since many of the cells that died early
were washed away. Thus, cyclic stretch caused apoptosis of transformed cells from several
different tissues (Figure 4C).
To determine if cyclic stretching conditions affected normal cells, the apoptosis assay was
performed on rigidity-dependent cells (including TPM2.1 restored cancer cells). When MDA-
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
11
MB-231-TPM2.1 cells were stretched, there was negligible apoptosis on rigid (6%) or soft (10%)
surfaces (Figure 4A and 4B). However, a dramatic increase in apoptosis (21%) was observed in
non-stretched MDA-MB-231-TPM2.1 cells on soft surfaces, showing that TPM2.1 restored
cancer cells behaved as normal rigidity-dependent cells. To further reconfirm these results,
similar experiments were performed on normal cells (MEFs and MCF10As) as the control cells.
Both stretched and non-stretched normal cells revealed negligible apoptosis (2-7%) on rigid
surfaces (Figure 4A, 4B and S4D). However, on soft surfaces, non-stretched control cells
experienced a substantial increase in apoptosis (41-45%), whereas cyclic stretching of the soft
surface protected them from apoptosis (5-10%). Thus, cyclic stretching of rigidity-dependent
cells inhibited apoptosis, particularly on soft surfaces, whereas cyclic stretching of their
transformed counterparts increased apoptosis on soft surfaces.
Cyclic Stretch Mediated-Cancer Cell Apoptosis Is Not Driven by DAPK1 Activation
Recently, we determined that tumor suppressor Death Associated Protein Kinase1
(DAPK1) causes rigidity-dependent cell apoptosis (anoikis) on soft surfaces but accelerates
adhesion assembly on rigid surfaces (Qin et al., 2018). To determine if DAPK1 was involved in
stretch-dependent cancer cell apoptosis, we blocked DAPK1 activity with a DAPK1 inhibitor
(DAPK1 inhibitor, 100 nM). DAPK1 inhibition decreased MEF apoptosis (14%) on non-
stretched soft surfaces (Figure 4D and Figure S7A). As previously reported, we also observed
distinct colocalization of DAPK1 with peripheral adhesions not only in stretched and non-
stretched normal cells (HFFs) on rigid surface but also in the cells on soft surfaces after cyclic
stretch (Figure S7B). In contrast, DAPK1 was cytoplasmic in normal cells on non-stretched soft
surfaces. When DAPK1 inhibitor was added to MDA-MB-231 cells during cyclic stretch, there
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
12
was no decline in apoptosis rate on rigid or soft surfaces indicating that a DAPK1-independent
apoptotic pathway was involved in cyclic stretch-induced transformed cancer cell apoptosis.
Calpain is the Major Downstream Effector of Stretch-Induced Cancer Cell Apoptosis
Since DAPK1 was not involved in cyclic stretch-induced cancer cell apoptosis, we looked
for another apoptotic mechanism and found that calpain proteases were implicated in inducing
apoptosis in cancer cells (Storr et al., 2011). Calpains were generally activated by a rise in
intracellular calcium level and then triggered apoptosis via caspase activity. To test the potential
role of calpains in cyclic stretch-induced apoptosis, ubiquitous calpain (calpain 1 & 2) activity
was inhibited using the calpain inhibitor (ALLN). Addition of ALLN inhibited apoptosis of
MDA-MB-231 cells after cyclic stretch compared to the non-treated counterpart (Figure 5A and
S8). Next, to assess the role of specific calpains in cancer cell apoptosis, calpain knock down
cells were developed using specific siRNAs. When cyclic stretch experiments were performed
on calpain 1 and calpain 2 KD MDA-MB-231 cells, both calpain 1 and 2 KD cells had lower
levels of apoptosis than control MDA-MB-231 cells (Figure 5A and S8). However, calpain 2 KD
cells showed significantly lower apoptosis, indicating that calpain 2 played the major role in
triggering cyclic stretch-induced cancer cell apoptosis (Figure 5B).
To understand the downstream effector of calpain cleavage, we focused on the pro-
apoptotic molecule BAX which can be cleaved by calpains to induce stress-mediated apoptosis
(Wood et al., 1998). Upon activation, BAX molecules translocated from cytoplasm to the
mitochondria to initiate the mitochondrial apoptotic pathway (Wolter et al., 1997). To check if
BAX was a downstream effector of calpain, MDA-MB-231 cells were stretched in the presence
of the BAX inhibitor peptide V5 (blocks mitochondrial translocation of BAX). Interestingly,
BAX inhibition reduced cell apoptosis (~10%) in stretched MDA-MB-231 cells (Figure 5C).
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
13
Thus, it appeared that BAX acted downstream of the calpain protease to initiate the
mitochondrial apoptotic pathway.
Cyclic Stretch-Induced Calcium Influx is the Upstream Effector of Cancer Cell Apoptosis
If stretch-induced apoptosis of transformed cells was caused by calpain, then the cellular level of
calcium should increase upon cyclic stretch. Although normal cells have tightly regulated
calcium channel expression and regulation (Azimi et al., 2014), greater calcium entry in
transformed cells due to cellular stress could have caused cell death via apoptosis (Baig et al.,
2016). To determine first if calcium entry was important, we treated cells with the nonspecific
calcium channel blocker, Gadolinium, before the application of stretch (Bourne and Trifaro,
1982). Interestingly it blocked the apoptosis in stretched cancer cells (Figure 6A), indicating a
possible role of calcium influx as the upstream effector of apoptosis. To actually measure
calcium levels during stretch, different cell lines (MDA-MB-231, MDA-MB-231-TPM2.1 and
MCF10A) were transfected with a calcium indicator, GECO1 and subjected to the cyclic stretch.
In the case of MDA-MB-231 cells, GECO1 intensity increased threefold after 2 hrs of cyclic
stretching. (Figure 6B) However, there was no increase in GECO1 intensity in MDA-MB-231-
TPM2.1 and MCF10A cells after 2 hrs of cyclic stretching, indicating that stretch triggered-
calcium influx did not take place in rigidity dependent-cancer cells and normal cells. Thus, a
stretch-induced increase in calcium levels was only found in the transformed cancer cells.
Mechanosensitive Piezo1 Channels Promote Cyclic Stretch-Induced Cell Apoptosis
The linkage between stretch-induced apoptosis and calcium entry implied the involvement of
mechanosensitive calcium channels. Piezo channels were activated by mechanical perturbations
of cells and enabled calcium entry to trigger intracellular calcium dependent signaling in many
physiological processes (Coste et al., 2010). We therefore compared the level of apoptosis in
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
14
Piezo1 expressing HEK 293 cells (transformed human embryonic kidney cells) and Piezo1
knock out HEK 293 cells (piezo KO HEK 293) after cyclic stretch on rigid surfaces for 6 hrs.
Cyclic stretch caused an increase in the apoptosis of piezo expressing HEK 293 cells, while
negligible apoptosis was observed in piezo KO HEK 293 cells (Figure 7A). Transformed HEK
293 cells lack endogenous TPM2.1. When the apoptosis assay was performed on TPM2.1
restored HEK 293 cells either with or without Piezo1, negligible apoptosis was noticed in piezo
KO HEK 293 or Piezo1 expressing HEK 293 cells with TPM2.1 restored (Figure 7B and Figure
S9). Thus, we suggest that stretch-induced transformed cell apoptosis involved Piezo1 channels
as well as the transformed state.
To determine if the rise in cellular calcium level relied upon Piezo1 channels, cyclic stretch
was applied on GECO1 transfected Piezo1- and Piezo1 KO-HEK 293 cells and TPM2.1 restored
Piezo1 HEK 293 cells. Of these three different cells, only piezo1 expressing HEK 293 cells
showed a twofold increase in GECO1 intensity within 30 min on rigid surface, while rigidity-
dependent or Piezo1 KO cells did not display any significant increase in the intensity (Figure
7C). All together, these findings confirmed that stretch-induced calcium entry and subsequent
apoptosis in transformed HEK 293 cells depended upon Piezo1 channels.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
15
Discussion
These results indicate that transformed cells exhibit cyclic stretch-induced apoptosis
regardless of the tissue origin, whereas rigidity-dependent normal cells show cyclic stretch-
induced growth and survival. For several different cell lines, the presence of TPM2.1 causes
rigidity-dependent growth, while depletion of TPM2.1 results in transformed growth despite
different tissue backgrounds. In transformed cells, cyclic stretch initially causes cell elongation
on rigid surfaces that is dependent on the stretch frequency as well as the magnitude of strain.
The frequency and magnitude of stretch were in a physiologically relevant range for exercise and
those parameters were used in all further studies. Longer period of cyclic stretch inhibits
transformed cell growth and increases apoptosis, particularly on soft surfaces. In contrast, cyclic
stretch promotes normal cell growth and inhibits apoptosis, particularly on soft surfaces. We find
that this relationship holds for cells from breast, ovaries, kidneys, connective tissue and skin.
Further, the stretch-induced apoptosis is calpain dependent (primarily calpain 2) and calpain acts
downstream of stretch-mediated calcium influx. The mechanosensitive Piezo1 calcium channels
are needed for calcium influx in the background of the transformed cell state.
Numerous studies regarding the transformed cell state imply that normal cells from a
variety of different tissues can be transformed. During the cell transformation, there is a shift
from an epithelial to mesenchymal phenotype in many cases; but the transformed phenotype is
strongly correlated with the loss of rigidity sensing and a change in the cortical actin
organization that makes cancer cell softer (Lin et al., 2015; Wolfenson et al., 2016). However,
the traction forces that most cancer cells exert on matrices are higher than their normal
counterparts and restoration of rigidity sensing in those cancer cells causes a decrease in traction
forces (Northcott et al., 2018; Wolfenson et al., 2016; Yang et al., 2018). Thus, it is logical to
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
16
consider the transformed state as a distinct cell state that can occur in cells from widely
different tissues with different expression patterns. Further, cells from different tissues can be
toggled between the transformed and the normal state by changing the level of expression of
several cytoskeletal proteins that enable or block the matrix rigidity sensing (Yang et al., 2018).
A relatively common feature of transformed cancer cells is that they have elevated levels
of calcium channels and calpain protease (Azimi et al., 2014; Storr et al., 2011). The higher level
of a mechanosensitive calcium channels and their altered functioning could explain the increased
susceptibility of transformed cancer cells to stretch-induced apoptosis. However, the
transformation of fibroblasts and epithelial cells with TPM2.1 KD will also sensitize those cells
to stretch-induced apoptosis. This indicates that transformed cells in general, whether from a
tumor or from a normal cell background, will undergo apoptosis in response to the appropriate
mechanical forces that can be mimicked by cyclic stretch.
There are previous reports of mechanical force-induced cancer cell death. Continuous
flow forces can cause the apoptosis of several different cancer cells attached to surfaces, whereas
oscillatory flow forces do not (Lien et al., 2013). The apoptotic pathway involves bone
morphogenetic protein receptor, Smad1/5, and p38 MAPK unlike our findings. Another study
shows that high shear forces on suspended circulating tumor cells initiates apoptosis possibly
through an oxidative stress-induced mitochondrial apoptotic pathway (Regmi et al., 2017).
Further, recent findings reveal that the stretching of mice for ten minutes a day for four weeks
suppresses the breast cancer growth by 50% compared to the non-stretched controls (Berrueta et
al., 2018). From these and other studies, there are strong indications that transformed cancer cells
may be mechanically vulnerable and thus sensitive to mechanical forces. Here we show that the
increased mechanical sensitivity is linked to the transformed cell state and not to a specific organ
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
17
or a specific cell type. Thus, it seems that this feature could be exploited to damage many
different types of tumor cells, particularly in a metastatic state where cells are not protected by
tumor fibrosis.
In Figure 8, we present a working model of mechanical stretch-induced transformed
cancer cell death, which we are describing as mechanoptosis. Prolonged application of
mechanical stresses on the transformed cells triggers a rise in calcium levels through altered
calcium homeostatic mechanisms, most likely increased Piezo1 activity. The rise in intracellular
calcium level causes activation of calpain 2 protease which initiates a mitochondrial apoptotic
pathway through its downstream effector, the BAX molecule. In contrast, mechanical stretching
of normal cells does not cause a rise in calcium levels because of a more tightly regulated
calcium homeostatic mechanism. Thus, cyclic stretch promotes the normal cell growth and
survival especially on the soft surfaces.
Although most cancer cells are transformed, there are many other aspects that are
important for tumor growth (Hanahan and Weinberg, 2000; Hirashima et al., 2013; Nishida et
al., 2006). Nevertheless, the transformed cell state appears to be associated with changes in
important cell functions such as the loss of rigidity sensing. Many other factors are also altered
beyond the loss of rigidity sensing in transformed cells, making it difficult to link the increased
mechanical sensitivity to a single factor. Despite the fact that we do not fully understand why
transformed cells are more sensitive to calcium loading with cyclic stretch, that sensitivity can
provide new approaches for killing cancer cells while actually stimulating the normal cell
growth. Hence, much more work is needed to determine if this feature can be exploited in an in
vivo setting to damage transformed cancer cells.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
18
Material and Methods:
Cell Culture
All cell lines MDA-MB-231(gift from Dr. Jay Groves, MBI, NUS), SKOV3 (gift from Dr. Ruby
Huang, CSI, NUS), HT1080 (ATCC), human foreskin fibroblasts (HFF, ATCC), transformed
human embryonic kidney cells (piezo expressing HEK 293 and piezo knock-out HEK 293, gift
from Boris Martinac, VCCRI Sydney and Ardem Patapoutian) and mouse embryonic fibroblasts
(MEF, ATCC) were cultured in high glucose DMEM (Thermo Fisher Scientific) supplemented
with 10% FBS and 1% penicillin and streptomycin at 37°C in 5% CO2 environment. MCF10A,
human breast epithelial cells (ATCC) were cultured in specialized growth medium as mentioned
before (Wolfenson et al., 2016). For cell experiments, high-glucose DMEM containing 10% FBS
and 1% penicillin and streptomycin was used. Cells were trypsinized using TrypLE (Thermo
Fisher Scientific) and cultured on human plasma fibronectin (10 μg/ml, Roche) or rat tail
collagen-I (20 μg/ml, Sigma) coated PDMS surface. Pharmacological inhibitors used are as
follows: DAPK1 inhibitor (100 nM, EMD Millipore), calpain inhibitor ALLN (100 μM, Sigma
Aldrich), non-specific calcium channel blocker gadolinium chloride (20 μM, Sigma Aldrich)
BAX inhibitor peptide V5 (200 μM, Merck). For inhibition assays, inhibitors were added to the
medium one hour post cell seeding and cyclic stretching was applied one to two hours after
addition of inhibitors. In general, cyclic stretching was applied 20 min after the cell seeding on
the cell stretching post (unless otherwise stated).
Plasmids and Transfection
For cell transfection using GFP tagged-TPM2.1 plasmid and RFP tagged-GECO1 calcium
indicator, Neon electroporation system (Invitrogen) was used according to the manufacturer’s
protocol. To perform TPM2.1 knockdown, cells were plated in a six well plate. TPM2.1 siRNA
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
19
(Qiagen) transfection was performed on the next day using lipofectamine RNAiMAX
(Invitrogen) according to the manufacturer’s protocol. Calpain siRNAs were prepared by Protein
Cloning Expression Facility, MBI, NUS using following siRNA sequences adapted from
previous reports. The calpain siRNA primers were: Calpain 1 primer (5’-AAGAC
CUAUGGCAUCAAGUGG-3’; 5’-AAGAAGUUGGUGAAGGGCCAU-3; ’5’-AAGCUAGU
GUUCGUGCACUCU-3’; 5’-AAG AGGAGAUUGACGAGAACU-3’ (Upla et al., 2008; Wu et
al., 2006) and Calpain 2 primer (5’-AAGACUUCACCGGAGGCAUUG-3’; 5’-AAGAUG
GGCGAGGACAUGCAC-3’) (Ma et al., 2012; Xu et al., 2010). Calpain knockdown cells were
generated by cell transfection using lipofectamine RNAiMAX.
Pillar and Flat Surface Fabrication
Pillar substrates were prepared with PDMS (10:1 prepolymer to curing agent ratio, Sylgard 184;
Dow Corning) spin coating on plastic molds at 1800 rpm for 1 min and cured at 80°C for 3 hrs.
The PDMS membrane was then peeled off the mold to serve as the surface for cell culture as
shown in Figure S1A. Pillars were fashioned in a square grid with 0.5 μm diameter, 1.8 μm
(~8kPa, soft) or 0.8 μm (~55kPa, rigid) height and 1.5 μm centre-to-centre distance among
pillars (Lohner et al., 2018). The flat PDMS layer (2 MPa rigid, 60-70 μm thick) was prepared by
spin coating PDMS (10:1 prepolymer to curing agent ratio) on a plastic sheet at 1800 rpm for 1
min, cured at 80°C for 3 hrs and peeled off to treat as the loading surface for cell culture.
Stretching Device Fabrication and Operation
The scheme of stretching device fabrication was shown in Figure S1A. Standard lithography
technique was used to fabricate the PDMS stretching device. Briefly, PDMS solution (10:1
prepolymer to curing agent ratio) was spin coated on a silicon wafer containing microfeatures at
500 rpm for 30 sec and cured at 80°C for 3 hrs. PDMS layer embossed with microfeatures was
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
20
peeled off, fixed to the glass bottom dish and subsequently covered with another thin PDMS
layer (60-70 μm thick). Glycerol (90%) was used as a lubricant to avoid sticking of thin PDMS
layer to the lower cell loading post. A PDMS cube with hole in middle was mounted on the top
of thin PDMS layer as shown in Figure S1A. Finally, the PDMS cube was connected to the
pump assembly. As depicted in Figure S1B & S1C, negative pressure was generated in the
hollow ring surrounding cell loading post by a pump which pulled the thin PDMS membrane
down to create membrane stretching. Different negative pressures were applied to measure the
magnitude of the strain generated at each pressure and the percent strain was then characterized
by measuring the changes in centre to centre distance between the pillars (Figure S1D & S1E).
To control the frequency of cyclic stretch, the Labview program was used.
Proliferation Assay
Cells were serum starved overnight to synchronize the cell cycle prior to cell seeding. EdU (5-
ethynyl-2’-deoxyuridine) reagent kit (Invitrogen) was used according to the manufacturer’s
instruction. The proliferation assay was performed after 9 hrs.
Apoptosis Assay
To assess cell apoptosis, Annexin V-Alexa Fluor 488 or Annexin V-Alexa Fluor 594 conjugates
(Thermofisher Scientific) were used according to the manufacturer’s protocol. Cells were
incubated with Annexin V solution after the specified time mentioned in the results section.
Immunocytochemistry and Fluorescence Microscopy
Cell samples were fixed with 4% paraformaldehyde (Thermofisher Scientific) in PBS for 10 min
and then permeabilized with 0.5% Triton X-100 for 5 min. Bovine Serum Albumin solution (2%)
(BSA) was used as a blocking buffer and samples were treated with BSA for 1 hr. Samples were
then incubated with primary antibodies of rabbit polyclonal anti DAPK1 (Abcam, 1:200), mouse
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
21
monoclonal anti Paxillin (BD Bioscience, 1:400) and mouse monoclonal anti tropomyosin
(1:200, Merck) for 1 hr at 37°C followed by treatment with Alexa Fluor-488 and -594 secondary
antibodies (Invitrogen). For F-actin staining, cells were incubated with Alexa Fluor 594
phalloidin antibody (Invitrogen) for 1 hr. DAPI dye (Merck) was used to stain cell nucleus.
Fluorescence and bright field images were acquired using Delta Vision System (Applied
Precision) on an Olympus IX70 microscope and equipped with CoolSNAP HQ2 CCD camera
(Photometrics, Tucson, USA). Live cell imaging was done using same Delta Vision System
maintained at 37°C, 5% CO2 condition. Optical images were acquired with an EVOS digital
fluorescence microscope (Fisher Scientific).
Western Blot
The cells were harvested after at least 48 hrs with the respective siRNA’s, pelleted, washed with
PBS and then lysed with RIPA buffer (Sigma-Aldrich) supplemented with the 1X cOmplete
Protease Inhibitor cocktail tablet (Roche, Cat no: 4693116001). The lysed cell mixture was
centrifuged at 15000 rpm, 4°C for 30 mins. The supernatant was collected and mixed with 2X
loading dye (2X Lammeli Buffer, Bio-Rad, Cat no: 1610737) + Beta-Mercaptoethanol (Sigma-
Aldrich). The samples were then denatured for 10 mins at 95°C and run on a 4-20% Mini-
PROTEAN ® TGX precast protein gels (Bio-Rad, Cat no: 4561094). The gels were then
transferred onto membranes. The membranes were blocked with 5% BSA solution in 1X TBST
(Tris-Buffered Saline with Tween-20) for 1 hr and incubated with primary antibody overnight at
4°C. The membranes were washed three times in TBST (10 mins per wash) followed by
secondary antibodies in 1XTBST (Horse Radish Peroxidase HRP) treatment for 1 hr. The
membranes were again washed three times in 1X TBST. The chemiluminescence of the
membranes was developed using Super Signal Femto Substrate Kit (Pierce) and developed using
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
22
ChemiDoc Touch Imager (Bio-Rad). Primary antibodies and conditions used are as follows:
Calpain 1 antibody (Rabbit, 1:1000, Abcam ab39170), Calpain 2 antibody (Rabbit, 1:1000,
Abcam ab39165), TPM 2.1 (Mouse, 1:1000, Merck), GAPDH (Mouse, 1:3000, Abcam ab8245).
The secondary antibodies, HRP-conjugated goat anti-rabbit IgG (bio-Rad 170-6515) and goat
anti-mouse IgG (bio-Rad 170-6516) were used at half of the primary antibody concentration.
STATISTICAL ANALYSIS
Statistical difference was calculated using two-tailed Student t-test. P value <0.05 was
considered statistically significant. All data represented as mean ± standard deviation.
ACKNOWLEDGEMENTS
We would like to thank all group members of Sheetz lab. A.T and M. Y are supported by the
Singapore Ministry of Education Academic Research Fund Tier 3 (MOE grant No: 2016 T3-1-
002). M.S is supported by NIH grants, NUS grants and Mechanobiology Institute, National
University of Singapore.
AUTHOR CONTRIBUTION
M.S and A.T designed study; A.T performed cell experiments, device fabrication and data
analysis; A.T and M.Y performed calcium indicator experiments; Y.W fabricated molds; Y.N
and C.T.L engineered stretching device; A.H developed western blots; A.T and M.S wrote
manuscript. All authors read manuscript and commented on it.
CONFLICT OF INTEREST STATEMENT
The authors declare no competing financial interest.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
23
Figure Legends
Figure 1. Effects of Magnitude and Frequency of Cyclic Stretching on Cancer Cell
Morphology with Different Matrix Rigidities
(A) MDA-MB-231 cells displayed the highest elongation (AR ~12) at 5% cyclic strain on flat
PDMS surfaces (rigid) after 6 hrs. (B) Cancer cell elongation was also found to be dependent on
the frequency of stretching, demonstrating maximum elongation at 0.5Hz with 5% cyclic strain
(AR ~11). (C) Matrix rigidity influenced the cancer cell elongation. Representative images
illustrating cyclic stretch-dependent cancer cell elongation on rigid surfaces (rigid pillars and flat
PDMS) but not on the soft pillar surfaces. (D) Statistical data showing marked increases in the
cell AR on rigid surfaces only. However, a stretch-dependent cell area increase was observed on
both rigid and soft surfaces after cyclic stretch. (E) Schematic diagram illustrating matrix
rigidity-dependent cancer cell elongation upon cyclic stretch. n=100 cells for cell aspect ratio and
n=30 cells for cell area. Experiments were repeated at least three times. *P< 0.05, **P< 0.01,
***P< 0.001. Scale bar: 20 µm.
Figure 2. TPM2.1 Restored Cancer Cells Spread upon Cyclic Stretch
(A) TPM2.1 restored MDA-MB-231 cells exhibited less elongation (AR~3) and increase in
spread area after 6 hrs of cyclic stretch in contrast to wild-type MDA-MB-231, (n=30 cells). (B)
SKOV3 cells also showed an elongated morphology with 6 hrs of cyclic stretch (AR~7)
compared to non-stretched cells (AR~3). No significant increment in the cell area was noticed
after cyclic stretching, (n=30 cells). (C) TPM2.1 restored SKOV3 cells displayed less elongation
(AR~2) and increase in spread area after cyclic stretch, (n=30 cells). (D) Representative images
showing the morphology of wild-type- and TPM2.1 KD-MEFs on a rigid surface after 6 hrs with
and without cyclic stretch. (E) Western blots showing the knockdown of TPM2.1 in MEFs. (F)
Wild-type MEFs showed spread morphology (AR~2) in response to cyclic stretch after 6 hrs,
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
24
while TPM2.1 KD MEFs were elongated (AR~4). The spread cell area increased for both wild-
type- and TPM 2.1KD-MEFs upon cyclic stretch, n=100 cells for aspect ratio and n=30 cells for
cell area measurement. (G) Schematic diagram showing the differential effect of cyclic stretch
on the morphology of transformed and normal cells. Experiments were repeated at least three
times. *P< 0.05, **P< 0.01, ***P< 0.001. Scale bar: 20 µm.
Figure 3. Cyclic Stretch Inhibits Transformed Cell Proliferation
(A) Time line depicting the cell proliferation assay protocol. (B) Representative images showing
the effect of cyclic stretch on the growth of transformed and normal cells (growing cells were
labelled by the anti-EdU antibody). (C) Cyclic stretch inhibited proliferation of M-231 breast
cancer cells on rigid and soft surfaces, while in TPM2.1 restored M-231 cells cyclic stretch
increased proliferation. Similarly, cyclic stretching of MEFs (control cells) increased
proliferation, whereas cyclic stretch of TPM2.1 KD MEF inhibited cell proliferation, (n>1000
cells for M-231, MEFs, TPM2.1 KD MEFs and n>500 cells for TPM2.1 restored M-231). (D)
Cyclic stretching of HT1080 fibrosarcoma cells also inhibited proliferation on rigid and soft
surfaces, (n>1000 cells). (E) Illustration depicting the cyclic stretch-dependent decline in
transformed cell growth and increase in normal cell growth. Experiments were repeated at least
two times. *P< 0.05, **P< 0.01, ***P< 0.001. Scale bar: 100 µm.
Figure 4. Cyclic Stretch Potentiates Transformed Cell Apoptosis
(A) Representative images illustrating the effect of 24 hrs cyclic stretch on apoptosis in M-231,
TPM2.1 restored M-231, MEFs and TPM2.1 KD MEFs. Dotted white line indicates location of
MEFs. (B) Cyclic stretch promoted M-231 cell apoptosis on rigid and soft surfaces. However,
TPM2.1 restored M-231 cells experienced negligible apoptosis upon cyclic stretch. Likewise,
MEFs (control cells) also demonstrated insignificant apoptosis with cyclic stretch. However,
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
25
non-stretched MEFs on soft surfaces displayed considerably high apoptosis similar to the non-
stretched TPM2.1 restored M-231 cells, indicating cell anoikis on the soft surface. In contrast,
TPM2.1 KD MEFs behaved like M-231 cells and experienced an increase in apoptosis after
cyclic stretch, (n>350 cells). (C) Diagram showing the effect of cyclic stretch on transformed
and normal cell apoptosis. (D) M-231 cells experienced notably high apoptosis upon stretch
even in the presence of DAPK1 inhibitor, suggesting the involvement of DAPK independent
apoptotic pathway. However, DAPK1 inhibition rescued the apoptosis in non-stretched MEFs on
soft surfaces showing that DAPK1 had a role in normal cell anoikis on soft surface, (n>300
cells). Data are representative of two independent experiments. *P< 0.05, **P< 0.01, ***P<
0.001. Scale bar: 50 µm.
Figure 5. Calpain Acts Downstream of Calcium to Initiate Cyclic Stretch-Induced
Apoptosis
(A) Cyclic stretch induced M-231 cell apoptosis decreased with Calpain inhibition by using
either a specific siRNA against Calpain-1 or -2, or a chemical inhibitor (ALLN). Specifically,
Calpain 2 KD caused greater inhibition of cell apoptosis compared to Calpain 1 KD, (n>1000
cells each for Calpain-1 and -2 knockdown assay and n>500 cell for ALLN inhibition assay and
control cells). (B) Schematic diagram highlighting the major role of Calpain 2 in promoting the
cyclic stretch-induced M-231 cell apoptosis. (C) M-231 cell apoptosis was further reduced by
BAX protein inhibition, implying that BAX protein acts downstream of the calpain protease,
(n>300 cells for BAX inhibition assay and n>600 cells for the control). Data are representative
of two independent experiments. **P< 0.01, ***P< 0.001. Scale bars: 50 µm.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
26
Figure 6. Cyclic Stretch-Mediated Calcium Influx is Upstream Activator of Transformed
Cell Apoptosis
(A) Cyclic stretch-dependent M-231 cell apoptosis was down regulated by non specific calcium
channel blocker, Gadolinium ion, (n>500 cells). Data are representative of two independent
experiments. ***P< 0.001. Scale bar: 50 µm. (B) Cells expressing GECO1 calcium indicator
were imaged on a cyclically stretched rigid surface. Representative time-lapse montages
displaying the progressive increase in GECO1 intensity in M-231 cells. In stark contrast, no such
increase in GECO1 intensity was observed in TPM2.1 restored M-231 cells and normal
MCF10A cells. After 2 hrs, the GECO1 intensity increased by three-fold in M-231 cells
compared to other groups, n>15 cells from three independent experiments were used. Scale bar:
50 µm.
Figure 7. Mechanosensitive Piezo1 Channels Needed for Cyclic Stretch-Induced Apoptosis
(A) Representative images illustrating elevated level of apoptosis in piezo expressing
transformed HEK293 cells upon cyclic stretch compared to piezo KO HEK 293 cells. (B) Piezo
expressing HEK293 cells exhibited significant apoptosis (~15%) upon cyclic stretch compared to
piezo KO HEK 293 and TPM2.1 restored HEK 293 either with or without Piezo1 (~4% in all
cases), (n>300 cells). Data are representative of two independent experiments. ***P< 0.001.
Scale bar: 50 µm. (C) Cells expressing GECO1 calcium indicator were monitored on a cyclically
stretched rigid surface. Representative time-lapse montages show a progressive increase in
GECO1 intensity in Piezo expressing HEK293 cells. No such increase in GECO1 intensity was
observed in TPM2.1 restored HEK293 and Piezo KO HEK293 cells. Two-fold increase in the
GECO1 intensity was observed in Piezo expressing HEK293 cells after 30 min compared to
other groups, n=16 cells from three independent experiments were used. Scale bar: 50 µm.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
27
Figure 8. Proposed Model of Transformed Cell Mechanoptosis
Prolonged cyclic stretching of transformed cells leads to intracellular calcium overloading,
activation of calpain protease and in turn initiation of the mitochondrial apoptotic pathway
through calpain downstream effector, BAX molecule. In contrast, mechanical stretching of
normal cells does not trigger calcium influx through their tightly regulated calcium channels. In
fact, it promotes the normal cell growth and survival through normal calcium homeostasis.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
28
References
Azimi, I., Roberts-Thomson, S.J., and Monteith, G.R. (2014). Calcium influx pathways in breast cancer:
opportunities for pharmacological intervention. British journal of pharmacology 171, 945-960.
Baig, S., Seevasant, I., Mohamad, J., Mukheem, A., Huri, H.Z., and Kamarul, T. (2016). Potential of
apoptotic pathway-targeted cancer therapeutic research: Where do we stand? Cell death & disease 7,
e2058.
Berrueta, L., Bergholz, J., Munoz, D., Muskaj, I., Badger, G.J., Shukla, A., Kim, H.J., Zhao, J.J., and
Langevin, H.M. (2018). Stretching Reduces Tumor Growth in a Mouse Breast Cancer Model. Scientific
reports 8, 7864.
Betof, A.S., Lascola, C.D., Weitzel, D., Landon, C., Scarbrough, P.M., Devi, G.R., Palmer, G., Jones,
L.W., and Dewhirst, M.W. (2015). Modulation of murine breast tumor vascularity, hypoxia and
chemotherapeutic response by exercise. Journal of the National Cancer Institute 107.
Bourne, G.W., and Trifaro, J.M. (1982). The gadolinium ion: a potent blocker of calcium channels and
catecholamine release from cultured chromaffin cells. Neuroscience 7, 1615-1622.
Coste, B., Mathur, J., Schmidt, M., Earley, T.J., Ranade, S., Petrus, M.J., Dubin, A.E., and Patapoutian,
A. (2010). Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels.
Science 330, 55-60.
Cui, Y., Hameed, F.M., Yang, B., Lee, K., Pan, C.Q., Park, S., and Sheetz, M. (2015). Cyclic stretching
of soft substrates induces spreading and growth. Nature communications 6, 6333.
Gateva, G., Kremneva, E., Reindl, T., Kotila, T., Kogan, K., Gressin, L., Gunning, P.W., Manstein, D.J.,
Michelot, A., and Lappalainen, P. (2017). Tropomyosin Isoforms Specify Functionally Distinct Actin
Filament Populations In Vitro. Current biology 27, 705-713.
Hamburger, A.W., and Salmon, S.E. (1977). Primary bioassay of human tumor stem cells. Science 197,
461-463.
Hanahan, D., and Weinberg, R.A. (2000). The hallmarks of cancer. Cell 100, 57-70.
Helfman, D.M., Flynn, P., Khan, P., and Saeed, A. (2008). Tropomyosin as a regulator of cancer cell
transformation. Advances in experimental medicine and biology 644, 124-131.
Hirashima, K., Migita, T., Sato, S., Muramatsu, Y., Ishikawa, Y., and Seimiya, H. (2013). Telomere
length influences cancer cell differentiation in vivo. Molecular and cellular biology 33, 2988-2995.
Kraning-Rush, C.M., Califano, J.P., and Reinhart-King, C.A. (2012). Cellular traction stresses increase
with increasing metastatic potential. PloS one 7, e32572.
Lien, S.C., Chang, S.F., Lee, P.L., Wei, S.Y., Chang, M.D., Chang, J.Y., and Chiu, J.J. (2013).
Mechanical regulation of cancer cell apoptosis and autophagy: roles of bone morphogenetic protein
receptor, Smad1/5, and p38 MAPK. Biochimica et biophysica acta 1833, 3124-3133.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
29
Lin, C.Y., Tseng, H.C., Shiu, H.R., Wu, M.F., Chou, C.Y., and Lin, W.L. (2012). Ultrasound sonication
with microbubbles disrupts blood vessels and enhances tumor treatments of anticancer nanodrug.
International journal of nanomedicine 7, 2143-2152.
Lin, H.H., Lin, H.K., Lin, I.H., Chiou, Y.W., Chen, H.W., Liu, C.Y., Harn, H.I., Chiu, W.T., Wang, Y.K.,
Shen, M.R., et al. (2015). Mechanical phenotype of cancer cells: cell softening and loss of stiffness
sensing. Oncotarget 6, 20946-20958.
Lohner, J., Rupprecht, J.-F., Hu, J., Mandriota, N., Saxena, M., Hone, J., Pitta de Araujo, D., Sahin, O.,
Prost, J., and Sheetz, M. (2018). Myosin filaments reversibly generate large forces in cells. bioRxiv.
Ma, J., Cui, W., He, S.M., Duan, Y.H., Heng, L.J., Wang, L., and Gao, G.D. (2012). Human U87
astrocytoma cell invasion induced by interaction of betaig-h3 with integrin alpha5beta1 involves calpain-
2. PloS one 7, e37297.
Marano, F., Frairia, R., Rinella, L., Argenziano, M., Bussolati, B., Grange, C., Mastrocola, R., Castellano,
I., Berta, L., Cavalli, R., et al. (2017). Combining doxorubicin-nanobubbles and shockwaves for
anaplastic thyroid cancer treatment: preclinical study in a xenograft mouse model. Endocrine-related
cancer 24, 275-286.
Nicolai, H., Steinbach, P., Knuechel-Clarke, R., Grimm, D., Roessler, W., Wieland, W.F., and
Hofstaedter, F. (1994). Proliferation of tumor spheroids after shock-wave treatment. Journal of cancer
research and clinical oncology 120, 438-441.
Nishida, N., Yano, H., Nishida, T., Kamura, T., and Kojiro, M. (2006). Angiogenesis in cancer. Vascular
health and risk management 2, 213-219.
Northcott, J.M., Dean, I.S., Mouw, J.K., and Weaver, V.M. (2018). Feeling Stress: The Mechanics of
Cancer Progression and Aggression. Frontiers in cell and developmental biology 6, 17.
Qin, R., Wolfenson, H., Saxena, M., and Sheetz, M. (2018). Tumor suppressor DAPK1 catalyzes
adhesion assembly on rigid but anoikis on soft matrices. bioRxiv.
Raval, G.N., Bharadwaj, S., Levine, E.A., Willingham, M.C., Geary, R.L., Kute, T., and Prasad, G.L.
(2003). Loss of expression of tropomyosin-1, a novel class II tumor suppressor that induces anoikis, in
primary breast tumors. Oncogene 22, 6194-6203.
Regmi, S., Fu, A., and Luo, K.Q. (2017). High Shear Stresses under Exercise Condition Destroy
Circulating Tumor Cells in a Microfluidic System. Scientific reports 7, 39975.
Stehn, J.R., Haass, N.K., Bonello, T., Desouza, M., Kottyan, G., Treutlein, H., Zeng, J., Nascimento,
P.R., Sequeira, V.B., Butler, T.L., et al. (2013). A novel class of anticancer compounds targets the actin
cytoskeleton in tumor cells. Cancer research 73, 5169-5182.
Storr, S.J., Carragher, N.O., Frame, M.C., Parr, T., and Martin, S.G. (2011). The calpain system and
cancer. Nature reviews Cancer 11, 364-374.
Swaminathan, V., Mythreye, K., O'Brien, E.T., Berchuck, A., Blobe, G.C., and Superfine, R. (2011).
Mechanical stiffness grades metastatic potential in patient tumor cells and in cancer cell lines. Cancer
research 71, 5075-5080.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
30
Upla, P., Marjomaki, V., Nissinen, L., Nylund, C., Waris, M., Hyypia, T., and Heino, J. (2008). Calpain 1
and 2 are required for RNA replication of echovirus 1. Journal of virology 82, 1581-1590.
Wolfenson, H., Meacci, G., Liu, S., Stachowiak, M.R., Iskratsch, T., Ghassemi, S., Roca-Cusachs, P.,
O'Shaughnessy, B., Hone, J., and Sheetz, M.P. (2016). Tropomyosin controls sarcomere-like contractions
for rigidity sensing and suppressing growth on soft matrices. Nature cell biology 18, 33-42.
Wolter, K.G., Hsu, Y.T., Smith, C.L., Nechushtan, A., Xi, X.G., and Youle, R.J. (1997). Movement of
Bax from the cytosol to mitochondria during apoptosis. The Journal of cell biology 139, 1281-1292.
Wood, D.E., Thomas, A., Devi, L.A., Berman, Y., Beavis, R.C., Reed, J.C., and Newcomb, E.W. (1998).
Bax cleavage is mediated by calpain during drug-induced apoptosis. Oncogene 17, 1069-1078.
Wu, M., Yu, Z., Fan, J., Caron, A., Whiteway, M., and Shen, S.H. (2006). Functional dissection of human
protease mu-calpain in cell migration using RNAi. FEBS letters 580, 3246-3256.
Xu, Y., Bismar, T.A., Su, J., Xu, B., Kristiansen, G., Varga, Z., Teng, L., Ingber, D.E., Mammoto, A.,
Kumar, R., et al. (2010). Filamin A regulates focal adhesion disassembly and suppresses breast cancer
cell migration and invasion. The Journal of experimental medicine 207, 2421-2437.
Yang, B., Lieu, Z.Z., Wolfenson, H., Hameed, F.M., Bershadsky, A.D., and Sheetz, M.P. (2016).
Mechanosensing Controlled Directly by Tyrosine Kinases. Nano letters 16, 5951-5961.
Yang, B., Wolfenson, H., Nakazawa, N., Liu, S., Hu, J., and Sheetz, M. (2018). Stopping Transformed
Growth with Cytoskeletal Proteins: Turning a Devil into an Angel. bioRxiv.
Zequi, S.C., Mourao, T.C., and Guimaraes, G.C. (2018). Prostate Cancer - Local Treatment after
Radiorecurrence: HIFU - High-Intensity Focused Ultrasound. International braz j urol : official journal of
the Brazilian Society of Urology 44, 429-432.
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
31
Figure 1
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
32
Figure 2
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
33
Figure 3
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
34
Figure 4
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
35
Figure 5
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
36
Figure 6
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
37
Figure 7
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
38
Figure 8
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/491746doi: bioRxiv preprint first posted online Dec. 10, 2018;
... Exploiting the differences between Ca 2+ signaling pathways of normal and malignant transformed cells may provide potentially viable strategies to specifically target cancer cells. We and others have shown previously that a fundamental difference between the transformed and normal phenotype is the response to cyclic stretch, which promotes the spreading and proliferation of normal cells but leads to Ca 2+ accumulation and apoptosis for transformed cells 2,25,26 . In the screening for the Ca 2+ channels responsible for cyclic-stretch dependent Ca 2+ entry in transformed cells, we determined that the mechanosensitive ion channel Piezo1 was essential 26 . ...
... We and others have shown previously that a fundamental difference between the transformed and normal phenotype is the response to cyclic stretch, which promotes the spreading and proliferation of normal cells but leads to Ca 2+ accumulation and apoptosis for transformed cells 2,25,26 . In the screening for the Ca 2+ channels responsible for cyclic-stretch dependent Ca 2+ entry in transformed cells, we determined that the mechanosensitive ion channel Piezo1 was essential 26 . Depletion of Piezo1 in transformed cell lines such as MDA-MB-231 and HEK 293T inhibited cyclicstretch dependent apoptosis. ...
... Depletion of Piezo1 in transformed cell lines such as MDA-MB-231 and HEK 293T inhibited cyclicstretch dependent apoptosis. Interestingly, in normal cells such as human foreskin fibroblast with comparable Piezo1 expression levels, cyclic stretching did not induce Ca 2+ build-up and apoptosis 26 . These results imply an important regulatory role for Piezo1 in Ca 2+ signaling of normal vs transformed cells. ...
Preprint
Full-text available
Piezo1 is a mechanosensitive Ca2+-permeable channel that has been implicated in a number of mechanosensing processes. It is diffusive on plasma membranes and activated by local membrane tension changes yet recent data suggest that Piezo1 activity is tightly coupled to the integrin-mediated actin cytoskeleton through a poorly understood mechanism. In our studies, we found that Piezo1 stably localizes to RGD matrix adhesions in normal but not in transformed cells. Piezo1 binding requires contractility and triggers local Ca2+ entry during adhesion formation and turnover. In transformed cells, Piezo1 level does not affect adhesion morphology; and there is no detectable Ca2+ influx around adhesions. Based upon these findings, we suggest that Piezo1 is a novel component of integrin-based adhesions in non-transformed cells and contributes to the distinct mechanosensing and Ca2+ signaling in normal vs transformed cells. Concentration of Piezo1 at adhesions is a key factor underlying Piezo1's physiological functions.
... Thus, phosphorylation of the lamina makes the NE more pliable and vulnerable to mechanical forces. Piezo channels are mechanosensitive ion channels that play key roles at the plasma membrane (56)(57)(58)(59). C. elegans has only one Piezo channel, PEZO-1 (60; 61). ...
Preprint
Full-text available
Successful sexual reproduction relies on robust quality control during meiosis. Assembly of the synaptonemal complex between homologous chromosomes (synapsis) regulates meiotic recombination and is crucial for accurate chromosome segregation in most eukaryotes. Synapsis defects can trigger cell cycle delays and, in some cases, apoptosis. Here, by developing and deploying a new chemically induced proximity system, we iden-tify key players in this quality control pathway in Caenorhabditis elegans . We find that persistence of the Polo-like kinase PLK-2 at pairing centers, specialized chromosome regions that interact with the nuclear envelope to promote homolog pairing and synapsis, induces apoptosis of oocytes by phosphorylating and destabilizing the nuclear lamina. Unexpectedly, we find that a mechanosensitive Piezo1/PEZO-1 channel localizes to the nuclear envelope and is required to transduce this signal to promote apoptosis. Thus, mechanosensitive ion channels play essential roles in detecting nuclear events and triggering apoptosis during gamete production. One-sentence summary Destabilization of the nuclear lamina triggers Piezo-dependent germline apoptosis.
Preprint
Full-text available
Cancer cells have many different behaviors from epithelial to mesenchymal forms. We report here that 36 distinct tumor cell lines regardless of EMT form or other features lack the ability to sense rigidity and will grow on soft surfaces. In the majority of lines, cells were missing at least one protein needed for rigidity sensing (primarily tropomyosin2.1 (Tpm2.1) but also PTPN12, FilaminA (FLNA), and myosinIIA) while all had high levels of Tpm3. In the few cases where the major rigidity sensing components were present, those tumor cells were not able to sense rigidity. Thus, we suggest that tumor cells can lose the ability to sense rigidity by many different means and that the loss of rigidity sensing is sufficient to cause the transformed phenotype that enables targeted treatments.
Article
Full-text available
There is growing interest in developing non-pharmacological treatments that could boost natural defenses against cancer and contribute to primary and secondary cancer prevention. Recent studies have shown that gentle daily stretching for 10 minutes can reduce local connective tissue inflammation and fibrosis. Because mechanical factors within the stroma can influence the tumor microenvironment, we hypothesized that stretching would reduce the growth of tumors implanted within locally stretched tissues and tested this hypothesis in a mouse orthotopic breast cancer model. Female FVB mice (N = 66) underwent bilateral injection of p53/PTEN double-null primary mouse mammary tumor cells into the third mammary fat pad. Mice were randomized to stretch vs. no stretch, and treated for 10 minutes once a day, for four weeks. Tumor volume at end-point was 52% smaller in the stretch group, compared to the no-stretch group (p < 0.001) in the absence of any other treatment. Cytotoxic immune responses were activated and levels of Specialized Pro-Resolving Mediators were elevated in the stretch group. These results suggest a link between immune exhaustion, inflammation resolution and tumor growth. Stretching is a gentle, non-pharmacological intervention that could become an important component of cancer treatment and prevention.
Article
Full-text available
The tumor microenvironment is a dynamic landscape in which the physical and mechanical properties evolve dramatically throughout cancer progression. These changes are driven by enhanced tumor cell contractility and expansion of the growing tumor mass, as well as through alterations to the material properties of the surrounding extracellular matrix (ECM). Consequently, tumor cells are exposed to a number of different mechanical inputs including cell–cell and cell-ECM tension, compression stress, interstitial fluid pressure and shear stress. Oncogenes engage signaling pathways that are activated in response to mechanical stress, thereby reworking the cell's intrinsic response to exogenous mechanical stimuli, enhancing intracellular tension via elevated actomyosin contraction, and influencing ECM stiffness and tissue morphology. In addition to altering their intracellular tension and remodeling the microenvironment, cells actively respond to these mechanical perturbations phenotypically through modification of gene expression. Herein, we present a description of the physical changes that promote tumor progression and aggression, discuss their interrelationship and highlight emerging therapeutic strategies to alleviate the mechanical stresses driving cancer to malignancy.
Article
Full-text available
Actin filaments assemble into a variety of networks to provide force for diverse cellular processes [1]. Tropomyosins are coiled-coil dimers that form head-to-tail polymers along actin filaments and regulate interactions of other proteins, including actin-depolymerizing factor (ADF)/cofilins and myosins, with actin [2–5]. In mammals, >40 tropomyosin isoforms can be generated through alternative splicing from four tropomyosin genes. Different isoforms display non-redundant functions and partially non-overlapping localization patterns, for example within the stress fiber network [6, 7]. Based on cell biological studies, it was thus proposed that tropomyosin isoforms may specify the functional properties of different actin filament populations [2]. To test this hypothesis, we analyzed the properties of actin filaments decorated by stress-fiber-associated tropomyosins (Tpm1.6, Tpm1.7, Tpm2.1, Tpm3.1, Tpm3.2, and Tpm4.2). These proteins bound F-actin with high affinity and competed with α-actinin for actin filament binding. Importantly, total internal reflection fluorescence (TIRF) microscopy of fluorescently tagged proteins revealed that most tropomyosin isoforms cannot co-polymerize with each other on actin filaments. These isoforms also bind actin with different dynamics, which correlate with their effects on actin-binding proteins. The long isoforms Tpm1.6 and Tpm1.7 displayed stable interactions with actin filaments and protected filaments from ADF/cofilin-mediated disassembly, but did not activate non-muscle myosin IIa (NMIIa). In contrast, the short isoforms Tpm3.1, Tpm3.2, and Tpm4.2 displayed rapid dynamics on actin filaments and stimulated the ATPase activity of NMIIa, but did not efficiently protect filaments from ADF/cofilin. Together, these data provide experimental evidence that tropomyosin isoforms segregate to different actin filaments and specify functional properties of distinct actin filament populations.
Article
Full-text available
Circulating tumor cells (CTCs) are the primary targets of cancer treatment as they cause distal metastasis. However, how CTCs response to exercise-induced high shear stress is largely unknown. To study the effects of hemodynamic microenvironment on CTCs, we designed a microfluidic circulatory system that produces exercise relevant shear stresses. We explore the effects of shear stresses on breast cancer cells with different metastatic abilities, cancer cells of ovarian, lung and leukemic origin. Three major findings were obtained. 1) High shear stress of 60 dynes/cm² achievable during intensive exercise killed more CTCs than low shear stress of 15 dynes/cm² present in human arteries at the resting state. 2) High shear stress caused necrosis in over 90% of CTCs within the first 4 h of circulation. More importantly, the CTCs that survived the first 4 h-circulation, underwent apoptosis during 16–24 h of post-circulation incubation. 3) Prolonged high shear stress treatment effectively reduced the viability of highly metastatic and drug resistant breast cancer cells. As high shear stress had much less damaging effects on leukemic cells mimicking the white blood cells, we propose that intensive exercise may be a good strategy for generating high shear stress that can destroy CTCs and prevent cancer metastasis.
Article
Full-text available
Underneath the intricacy of every cancer lies mysterious events that impel the tumour cell and its posterity into abnormal growth and tissue invasion. Oncogenic mutations disturb the regulatory circuits responsible for the governance of versatile cellular functions, permitting tumour cells to endure deregulated proliferation, resist to proapoptotic insults, invade and erode normal tissues and above all escape apoptosis. This disruption of apoptosis has been highly implicated in various malignancies and has been exploited as an anticancer strategy. Owing to the fact that apoptosis causes minimal inflammation and damage to the tissue, apoptotic cell death-based therapy has been the centre of attraction for the development of anticancer drugs. Increased understanding of the molecular pathways underlying apoptosis has enabled scientists to establish unique approaches targeting apoptosis pathways in cancer therapeutics. In this review, we reconnoitre the two major pathways (intrinsic and extrinsic) targeted cancer therapeutics, steering toward chief modulators of these pathways, such as B-cell lymphoma 2 protein family members (pro- and antiapoptotic), inhibitor of apoptosis proteins, and the foremost thespian of extrinsic pathway regulator, tumour necrosis factor-related apoptosis-inducing agent. Together, we also will have a look from clinical perspective to address the agents (drugs) and therapeutic strategies adopted to target these specific proteins/pathways that have entered clinical trials.
Article
Full-text available
Cells test the rigidity of the extracellular matrix by applying forces to it through integrin adhesions. Recent measurements show that these forces are applied by local micrometre-scale contractions, but how contraction force is regulated by rigidity is unknown. Here we performed high temporal- and spatial-resolution tracking of contractile forces by plating cells on sub-micrometre elastomeric pillars. We found that actomyosin-based sarcomere-like contractile units (CUs) simultaneously moved opposing pillars in net steps of ∼2.5 nm, independent of rigidity. What correlated with rigidity was the number of steps taken to reach a force level that activated recruitment of α-actinin to the CUs. When we removed actomyosin restriction by depleting tropomyosin 2.1, we observed larger steps and higher forces that resulted in aberrant rigidity sensing and growth of non-transformed cells on soft matrices. Thus, we conclude that tropomyosin 2.1 acts as a suppressor of growth on soft matrices by supporting proper rigidity sensing.
Article
Full-text available
The stiffness sensing ability is required to respond to the stiffness of the matrix. Here we determined whether normal cells and cancer cells display distinct mechanical phenotypes. Cancer cells were softer than their normal counterparts, regardless of the type of cancer (breast, bladder, cervix, pancreas, or Ha-RasV12-transformed cells). When cultured on matrices of varying stiffness, low stiffness decreased proliferation in normal cells, while cancer cells and transformed cells lost this response. Thus, cancer cells undergo a change in their mechanical phenotype that includes cell softening and loss of stiffness sensing. Caveolin-1, which is suppressed in many tumor cells and in oncogene-transformed cells, regulates the mechanical phenotype. Caveolin-1-upregulated RhoA activity and Y397FAK phosphorylation directed actin cap formation, which was positively correlated with cell elasticity and stiffness sensing in fibroblasts. Ha-RasV12-induced transformation and changes in the mechanical phenotypes were reversed by re-expression of caveolin-1 and mimicked by the suppression of caveolin-1 in normal fibroblasts. This is the first study to describe this novel role for caveolin-1, linking mechanical phenotype to cell transformation. Furthermore, mechanical characteristics may serve as biomarkers for cell transformation.
Article
Full-text available
Exercise has been shown to improve postischemia perfusion of normal tissues; we investigated whether these effects extend to solid tumors. Estrogen receptor-negative (ER-, 4T1) and ER+ (E0771) tumor cells were implanted orthotopically into syngeneic mice (BALB/c, N = 11-12 per group) randomly assigned to exercise or sedentary control. Tumor growth, perfusion, hypoxia, and components of the angiogenic and apoptotic cascades were assessed by MRI, immunohistochemistry, western blotting, and quantitative polymerase chain reaction and analyzed with one-way and repeated measures analysis of variance and linear regression. All statistical tests were two-sided. Exercise statistically significantly reduced tumor growth and was associated with a 1.4-fold increase in apoptosis (sedentary vs exercise: 1544 cells/mm(2), 95% CI = 1223 to 1865 vs 2168 cells/mm(2), 95% CI = 1620 to 2717; P = .048), increased microvessel density (P = .004), vessel maturity (P = .006) and perfusion, and reduced intratumoral hypoxia (P = .012), compared with sedentary controls. We also tested whether exercise could improve chemotherapy (cyclophosphamide) efficacy. Exercise plus chemotherapy prolonged growth delay compared with chemotherapy alone (P < .001) in the orthotopic 4T1 model (n = 17 per group). Exercise is a potential novel adjuvant treatment of breast cancer. © The Author 2015. Published by Oxford University Press.
Article
Anaplastic thyroid cancer is one of the most lethal diseases and a curative therapy does not exist. Doxorubicin, the only drug approved for anaplastic thyroid cancer treatment, has a very low response rate and causes numerous side effects among which cardiotoxicity is the most prominent. Thus, doxorubicin delivery to the tumor site could be an import goal aimed to improve drug efficacy and to reduce its systemic side effects. We recently reported that, in human anaplastic thyroid cancer cell lines, combining doxorubicin-loaded nanobubbles with extracorporeal shock waves, acoustic waves used in lithotripsy and orthopedics without side effects, increased the intracellular drug content and in vitro cytotoxicity. In the present study we tested the efficacy of this treatment on a human anaplastic thyroid cancer xenograft mouse model. After 21 days, the combined treatment determined the greatest drug accumulation in tumors with consequent reduction of tumor volume and weight, and an extension of the tumor doubling time. Mechanistically, the treatment induced tumor apoptosis and decreased cell proliferation. Finally, whereas doxorubicin caused the increase of fibrosis markers and oxidative stress in animal hearts, loading doxorubicin into nanobubbles avoided these effects preventing heart damage. The improvement of doxorubicin anti-tumor effects together with the prevention of heart damage suggests that the combination of doxorubicin-loaded nanobubbles with extracorporeal shock waves might be a promising drug delivery system for anaplastic thyroid cancer treatment.