Cytoskeletal role in differential adhesion patterns of normal fibroblasts and breast cancer cells inside silicon microenvironments.
ABSTRACT In this paper we studied differential adhesion of normal human fibroblast cells and human breast cancer cells to three dimensional (3-D) isotropic silicon microstructures and investigated whether cell cytoskeleton in healthy and diseased state results in differential adhesion. The 3-D silicon microstructures were formed by a single-mask single-isotropic-etch process. The interaction of these two cell lines with the presented microstructures was studied under static cell culture conditions. The results show that there is not a significant elongation of both cell types attached inside etched microstructures compared to flat surfaces. With respect to adhesion, the cancer cells adopt the curved shape of 3-D microenvironments while fibroblasts stretch to avoid the curved sidewalls. Treatment of fibroblast cells with cytochalasin D changed their adhesion, spreading and morphology and caused them act similar to cancer cells inside the 3-D microstructures. Statistical analysis confirmed that there is a significant alteration (P < 0.001) in fibroblast cell morphology and adhesion property after adding cytochalasin D. Adding cytochalasin D to cancer cells made these cells more rounded while there was not a significant alteration in their adhesion properties. The distinct geometry-dependent cell-surface interactions of fibroblasts and breast cancer cells are attributed to their different cytoskeletal structure; fibroblasts have an organized cytoskeletal structure and less deformable while cancer cells deform easily due to their impaired cytoskeleton. These 3-D silicon microstructures can be used as a tool to investigate cellular activities in a 3-D architecture and compare cytoskeletal properties of various cell lines.
- Journal of The American College of Cardiology - J AMER COLL CARDIOL. 01/2011; 57(14).
- [Show abstract] [Hide abstract]
ABSTRACT: The physiological role of the actin cytoskeleton is well known: it provides mechanical support and endogenous force generation for formation of a cell shape and for migration. Furthermore, a growing number of studies have demonstrated another significant role of the actin cytoskeleton: it offers dynamic epigenetic memory for guiding cell fate, in particular, proliferation and differentiation. Because instantaneous imbalance in the mechanical homeostasis is adjusted through actin remodeling, a synthetic extracellular matrix (ECM) niche as a source of topographical and mechanical cues is expected to be effective at modulation of the actin cytoskeleton. In this context, the synthetic ECM niche determines cell migration, proliferation, and differentiation, all of which have to be controlled in functional tissue engineering scaffolds to ensure proper regulation of tissue/organ formation, maintenance of tissue integrity and repair, and regeneration. Here, with an emphasis on the epigenetic role of the actin cytoskeletal system, we propose a design concept of micro-/nanotopography of a tissue engineering scaffold for control of cell migration, proliferation, and differentiation in a stable and well-defined manner, both in vitro and in vivo .Tissue Engineering Part B Reviews 04/2014; · 4.64 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: We report the fabrication, functionalization and testing of microdevices for cell culture and cell traction force measurements in three-dimensions (3D). The devices are composed of bent cantilevers patterned with cell-adhesive spots not lying on the same plane, and thus suspending cells in 3D. The cantilevers are soft enough to undergo micrometric deflections when cells pull on them, allowing cell forces to be measured by means of optical microscopy. Since individual cantilevers are mechanically independent of each other, cell traction forces are determined directly from cantilever deflections. This proves the potential of these new devices as a tool for the quantification of cell mechanics in a system with well-defined 3D geometry and mechanical properties.Lab on a Chip 11/2013; · 5.70 Impact Factor
Cytoskeletal role in differential adhesion patterns
of normal fibroblasts and breast cancer cells inside
Mehdi Nikkhah & Jeannine S. Strobl & Bhanu Peddi &
Published online: 17 December 2008
# Springer Science + Business Media, LLC 2008
Abstract In this paper we studied differential adhesion of
normal human fibroblast cells and human breast cancer
cells to three dimensional (3-D) isotropic silicon micro-
structures and investigated whether cell cytoskeleton in
healthy and diseased state results in differential adhesion.
The 3-D silicon microstructures were formed by a single-
mask single-isotropic-etch process. The interaction of these
two cell lines with the presented microstructures was
studied under static cell culture conditions. The results
show that there is not a significant elongation of both cell
types attached inside etched microstructures compared to
flat surfaces. With respect to adhesion, the cancer cells
adopt the curved shape of 3-D microenvironments while
fibroblasts stretch to avoid the curved sidewalls. Treatment
of fibroblast cells with cytochalasin D changed their
adhesion, spreading and morphology and caused them act
similar to cancer cells inside the 3-D microstructures.
Statistical analysis confirmed that there is a significant
alteration (P<0.001) in fibroblast cell morphology and
adhesion property after adding cytochalasin D. Adding
cytochalasin D to cancer cells made these cells more
rounded while there was not a significant alteration in their
adhesion properties. The distinct geometry-dependent cell–
surface interactions of fibroblasts and breast cancer cells are
attributed to their different cytoskeletal structure; fibroblasts
have an organized cytoskeletal structure and less deform-
able while cancer cells deform easily due to their impaired
cytoskeleton. These 3-D silicon microstructures can be used
as a tool to investigate cellular activities in a 3-D
architecture and compare cytoskeletal properties of various
Recent reports have emphasized that cellular functions are
controlled by their surrounding microenvironment (Semino
et al. 2003; Takezawa 2003) including substrate topogra-
phy, surface chemistry, soluble factors, and signals from
neighboring cells. Microengineered environments can fa-
cilitate quantitative analysis of the effects of microscale
surface topography on individual cell growth, adhesion,
migration, and mechanics. They can also be used to explore
how cell-cell interactions influence the aforementioned
cellular behaviors (Clark et al. 1987, 1990). Understanding
cell-surface interactions in vitro has broad applicability in
cell-based biosensors, tissue engineering, prosthetics, and
Biomed Microdevices (2009) 11:585–595
M. Nikkhah (*)
Department of Mechanical Engineering, Virginia Tech,
Blacksburg, VA 24061, USA
J. S. Strobl
Edward Via Virginia College of Osteopathic Medicine,
2265 Kraft Drive,
Blacksburg, VA 24060, USA
M. Nikkhah:B. Peddi:M. Agah
Virginia Tech MEMS Laboratory, The Bradley Department
of Electrical and Computer Engineering, Virginia Tech,
Blacksburg, VA 24061, USA
The influence of surface topography on cellular behavior
was first noted by Harrison (Harrison 1914) and led to the
concept of “contact guidance” in which cells align in the
direction of surface grooves. Advances in microelectrome-
chanical systems (MEMS) technology have made it
possible to observe cellular responses to a variety of precise
topographical patterns including microgrooves fabricated
from silicon (DenBraber et al. 1996), silicone (Walboomers
et al. 1999a, b), quartz (Clark et al. 1991; Dalby et al. 2003;
Rajnicek et al. 1997), titanium-coated silicon (Oakley and
Brunette 1993; Oakley et al. 1997; Walboomers et al.
1999a, b), polystyrene (Matsuzaka et al. 2000; Walboomers
et al. 1998, 1999a, b), silica (Wojciakstothard et al. 1995)
and ploy (glycerol-sebacate) (Bettinger et al. 2006), arrays
of pillars and wells (Rovensky et al. 1991; Turner et al.
2000; van Kooten and von Recum 1999), arrays of grooves
and holes fabricated in silicon (Charest et al. 2007) and
polystyrene (Dusseiller et al. 2005), arrays of pyramid
microstructures (Liao et al. 2003) fabricated in silicone and
networks of ridges and grids fabricated in PDMS (Mai et al.
2007). Some of the aforementioned studies have focused on
the critical role of cytoskeletal components, actin micro-
filaments and microtubules in cell alignment (Oakley et al.
1997; Wojciakstothard et al. 1995). Both mechanical
loading and microgrooved surface topographies promote
cytoskeletal reorganization and affect the behavior of
fibroblast cells (Loesberg et al. 2005, 2006). Cytoskeletal
proteins play an important role in tissue engineering by
mediating the cellular response to different substrates.
The devices described in the previously cited literatures
involve microstructures with vertical sidewalls. However,
extracellular matrix (ECM) proteins are well known to
form a complex interconnecting network and a three
dimensional surface topography (Timpl 1996). Previously,
we reported development of arrays of 3-D silicon microstruc-
tures consisting of isotropic microchambers interconnected
with channels in order to asses the adhesion and growth
behavior of normal human fibroblast cells (HS68) and human
breast cancer cells (MDA-MB-231) (Nikkhah et al. 2008).
These cells represent key cell types in human breast tumor
microenvironments (Kalluri and Zeisberg 2006). The main
focus in our previous work was the growth behavior of these
cell lines inside isotropic microstructures; however, we found
that these cells exhibited differential adhesion to the micro-
structures with curved sidewalls. In this work, we explore in
detail whether the differential adhesion of these two cell
types to isotropic architectures is dependent on their
cytoskeletal structure and biomechanical properties. For this
purpose, we developed new 3-D isotropic silicon micro-
structures using our novel fabrication technology (Gantz et
al. 2008). These microenvironments can be potentially used
as novel platforms to study cytoskeletal organization of
various cell lines in a 3-D architecture.
2 Materials and methods
2.1 Device fabrication
The final devices shown in Fig. 1 were fabricated from
single-crystal silicon wafers. The fabrication process is
similar to our previous work relying on the applicability of
reactive ion etching (RIE) lag and its dependence on
geometrical patterns of the photomask layout to etch silicon
to different depths (Nikkhah et al. 2008). Figure 2 shows
different layouts of the photomask design; the openings for
the star-shaped microchambers are rectangular with the
dimension in the range of 2 μm×2 μm to 10 μm×10 μm
(Fig. 2(a–b)). The openings for the circular-shaped micro-
chambers are circular rings with a diameter of 5 μm and
withvariablespacing ragingfrom2 μm to6 μm (Fig. 2(c–d)).
Briefly, the fabrication process starts by depositing 8,000 Å-
thick PECVD1oxide layer on a silicon wafer. After spinning
and patterning photoresist, the oxide layer was etched for
3 min using DRIE2CH4/C4F8plasma. Next, silicon was
etched using DRIE SF6plasma to form complex arrays of
features composed of star and circular-shaped microcham-
bers. Due to RIE lag, areas under the bigger mask openings
were etched more, thereby achieving 3-D microstructures.
After removing photoresist the oxide layer is subsequently
removed using DRIE.
Figure 3 shows the scanning electron microscopy
(SEM)3images of the microchambers with varying cross
sectional shapes. The depth of the microchambers varies
between 60 and 70 µm and the width ranges between
150 and 170 μm. Table 1 summarizes the etching
parameters. It is notable that the described fabrication
technique provides localized rough edges on the curved
sidewalls and on the bottom surface of the microcham-
bers. In the star patterns, the rough edges formed on the
sidewalls (Fig. 3(a)) while in the circular-shaped micro-
chambers, the rough edges were localized on the bottom
of the microchambers in the form of concentric rings
The final fabricated wafer is diced into 1 cm2chips, each
chip contains the etched microstructures surrounded by
un-etched flat surfaces (Fig. 1). This provides the ability to
compare the control experiments on flat surfaces and the
results on the etched microstructures under the same cell
culturing conditions. Chips were cleaned using acetone and
isopropanol, rinsed in deionized water, and, finally, air-
dried prior to cell culture experiments.
586Biomed Microdevices (2009) 11:585–595
2.2 Cell culture and reagents
HS68 normal human fibroblasts and MDA-MB-231 human
breast cancer cells were purchased from the American Type
Culture Collection (ATCC). Cells were maintained in
plastic T-75 cm2culture flasks in RPMI culture medium
which contained 10% fetal bovine serum (FBS), 1 mM
sodium pyruvate, and penicillin–streptomycin (100 U/ml).
Cells were grown at 37°C in a humidified 7% CO2–93% air
atmosphere. Cells were fixed in a 50% ethanol–5%
formaldehyde–0.1% crystal violet solution in normal saline
for 10 min prior to imaging. After 72 h of growth,
cytochalasin D (final concentration of 10 μM) (Kojic et
al. 2007) was added to the culture medium for either 10 or
30 min prior to fixation.
2.3 Microscopy and Immunofluorescence for cytoskeletal
Optical microscopy was performed using a digital-video
optical microscope4with a 2,000 maximum magnification
for cell counting on flat silicon and glass surfaces. Confocal
microscopy5was performed in the reflection mode to image
the actin cytoskeletal structures. Cell nuclei were stained
with Hoechst 33342 dye and actin was visualized using
Alexa Fluor-488 phalloidin (Turner et al. 2000). The
confocal images inside the microchambers were taken at
different depths of focus to construct Z-stack images. SEM
was also performed to assess the detailed effects of the
etched microstructures on the cells’ morphology. After
fixation of the culture with 3.7% formaldehyde in phos-
phate buffered saline (PBS) for 10 min, samples were
critical point dried and sputter-coated with a thin layer of
gold palladium prior to SEM imaging.
2.4 Quantification of cell number, cell length and cell
SEM images were obtained at nine different fields of star
and circular-shaped microchambers within each chip. Each
field included 24 microchambers and their adjacent flat
surfaces. The selected fields almost cover the entire
fabricated substrate. These images were used to count the
number of fibroblast cells inside the etched features before
and after adding Cytochalasin D.
Cell length measurements on flat surfaces were per-
formed using the National Institutes of Health (NIH) ImageJ
(v. 1.41) software. The cell length was quantified as the
distance between farthest end points of the cell. To quantify
the cell length inside etched microchambers, we assumed an
ellipsoid shape for the etched microchambers with an
equatorial radii and “c” refers to the polar radius (micro-
chamber depth). For the star-shaped microchambers, a=
80 μm, b=80 μm, c=55 μm, whereas for the circular-shaped
microchambers, a=82.5 μm, b=82.5 μm, c=70 μm. The x
and y coordinates of the end points of the cell were measured
c2¼ 1, where “a” and “b”are the
Fig. 2 Photo mask layout comprising four different features for
star (a, b) and circular-shaped microchambers (c, d). The openings
for the star-shaped microchambers are rectangular with the dimen-
sion in the range of 2 μm×2 μm to 10 μm×10 μm. The openings for
the circular-shaped microchambers are circular rings with a diameter
of 5 μm and with variable spacing ranging from 2 to 6 μm
Fig. 1 Photo image of the fabricated devices in silicon. The cell
culture substrate comprises star and circular-shaped microchambers
5ZEISS LSM 510 META
Biomed Microdevices (2009) 11:585–595 587
from the top view images using ImageJ software. Then the z
coordinate of the corresponding points was measured using
the ellipsoid equation stated above. Finally the cell length is
measured using the standard line equation in 3-D space
For fibroblast cell shape analysis, three different mor-
phologies of round, spread and stretched were assumed.
The spread cells are those in which the cell body has a
uniform contact with the sidewalls of the microchamber
while the stretched ones anchor to different location of the
microchambers and do not have contact with the sidewalls.
The average aspect ratio Max length=width
and spread cells was measured to be 26.6±13.8 and 3.7±
ðÞ2þ y2? y1
ðÞ2þ z2? z1
ðÞÞ of stretched
2.5 Statistical analysis
Two-way ANOVA analysis was performed using Prism v.56
to study the effect of etched geometries on cell length and
to assess the statistical significance of the effect of
cytochalasin D on the morphology of cells growing in
circular and star-shaped microchambers.
3.1 Fibroblast and cancer cell growth on flat surfaces
We cultured each cell type on flat silicon and on glass
surfaces as our control substrates. Following seeding at low
density (442 cells/mm2), fibroblast cells attached and
spread on both surfaces within 6 h while the corresponding
time for cancer cells was about 24 h. This process was
visualized under an optical microscope. HS68 cells prolif-
erated on silicon and glass surfaces with typical cell growth
kinetics. Figure 4(a) demonstrates logarithmic and plateau
phases of fibroblast cell growth. There was not a significant
difference in the growth of these cells on silicon compared
to glass—which is commonly used for biological experi-
ments. Figure 4(b) demonstrates the logarithmic phase of
the MDA-MB-231 cell growth on silicon and glass
surfaces. Cancer cells showed similar growth rates on both
6GraphPad Software, San Diego, CA, USA.
Fig. 3 SEM images of the three
dimensional (3-D) silicon sub-
strates comprising star and
with the corresponding top view
images. With the proposed fab-
rication technology, scalloped
edges can be formed on the
curved sidewalls and the bottom
surface of the microchambers
Table 1 Etching parameters for the isotropic geometries
Etching parameters3004.5e−021,800 10
588Biomed Microdevices (2009) 11:585–595
substrates. The logarithmic growth rate of breast cancer cells
typically exceeds that of fibroblast cells because transformed
cells bear oncogenic mutations (Weinberg 2007). Typically,
cancer cells achieved higher cell densities than fibroblast
cells. After 168 h in culture, confluent monolayers of cells
covered the smooth surfaces of the chip.
3.2 Fibroblast and cancer cell behavior inside the etched
Figure 5 compares the length of fibroblast and cancer cells
on flat surfaces as well as inside the etched microstructures.
An average length of 88±23 μm (n=100) and 33±11 μm
(n=100) was measured for fibroblast and cancer cells,
respectively. The geometry of the etched substrate had
insignificant effect on the length of either cell line. HS68
fibroblast cells are significantly longer (p<0.001) than the
MDA-MB-231 breast cancer cells on flat surfaces and all
the etched geometries. Both cell types distributed with
random orientation within the etched microstructures (data
Figure 6(A–F) shows the SEM images of untreated
control fibroblast cells inside the etched features revealing
the tendency of fibroblasts to stretch inside the isotropic
microchambers to avoid curved sidewalls. Cancer cells, on
the other hand, attached to the curved sidewalls of the
microchambers (Fig. 6(G–L)). The cross-sectional views
clearly demonstrate that cancer cells adapted their shape to
fit the curved sidewalls and their adhesion behavior was
less responsive to the substrate geometrical pattern com-
pared to that of normal fibroblast cells. This observation is
in agreement with our previous work which was carried out
on different isotropic microenvironments (Nikkhah et al.
To examine whether cell cytoskeletal structure plays a
major role in the adhesion behavior of these two cell lines
we disrupted the actin cytoskeleton with cytochalasin D.
Confocal microscopy was performed after 72 h of cell grown
on flat silicon substrates to visualize the disruption of the
actin cytoskeleton for different treatment times of 10 and
30 min and to show the dynamics of cell cytoskeleton
alteration. Figure 7 illustrates the actin cytoskeleton and
nuclei distribution in fibroblasts attached to flat silicon
surfaces. The actin filaments were distributed in long, thin
bundles with parallel or cross-link distribution (Fig. 7(A–
B)). After 10 min of treatment with cytochalasin D, there
were persistent actin filament bundles through the cell
body (Fig. 7(C–D)). However after 30 min of treatment,
the dense mesh and long fibers of actin were entirely
disrupted throughout the cell body (Fig. 7(E-F)).
Figure 8 shows the actin cytoskeleton of the MDA-MB-
231 cancer cells attached to flat silicon surfaces. The actin
cytoskeletal structure of the cancer cells was organized
differently compared to fibroblast cells. The cancer cells
showed no actin filament formation. The actin cytoskeleton
of cancer cells, attached to silicon was distributed in the
form of concentric rings around the nuclei. We repeated
these experiments five times and the results were highly
reproducible. The actin cytoskeleton of cancer cells was
Fig. 4 (a) Fibroblast and (b) cancer cells growth on flat silicon and
glass surfaces as a function of time showing typical cell growth
kinetics. Data shown are the mean and SD of the cells per mm2
Fig. 5 Fibroblast and cancer cell length on flat surfaces and inside the
star and circular-shaped microchamber
Biomed Microdevices (2009) 11:585–595589
completely disrupted after 10 min treatment with cytocha-
lasin D (Fig. 8(C–D)).
Figure 9 shows the SEM images of the fibroblast and
cancer cells attached inside the etched features after
treatment with cytochalasin D. After 10 min in the presence
of cytochalasin D, some fibroblast cells lost their stretching
behavior and well spread and adapted to the curved
sidewalls (Fig. 9(A–C)). After 30 min, however, the
majority of the cells lost their stretching behavior and
rounded up indicating that these cells lost almost all their
entire stress fibers consistent with the confocal microscopy
observations (Fig. 9(D–F)). At this state, the average length
of the rounded fibroblast cells excluding the dendritic
extensions was measured to be approximately 16±2 μm (n
=20). On the other hand, cancer cells became more rounded
after treatment with cytochalasin D and there was not a
significant change in their adhesion behavior (Fig. 9(G–I)).
Similar to their original state, they attached to the curved
sidewalls of the microchambers. At this state, the length of
these cells was measured to be approximately 17±2 μm
(n=20). Figure 10 shows the quantitative data regarding the
effect of cytochalasin D on fibroblasts for both 10 and
30 min treatment times. After 10 min disruption of actin
cytoskeleton, 35±7% of the cells significantly (P<0.001)
lost their anchorage dependent adhesion inside the etched
features and spread on the curved sidewalls. However,
30 min treatment with cytochalasin D significantly (P<
0.001) altered the morphology of 70±7% of fibroblast cells
toward rounded morphology associated with dendritic
extensions. In such state, their adhesion became more like
that of the cancer cells; that is adapting to the etched
microstructures and taking the shape of the curved sidewalls.
Fig. 6 SEM images of the HS68 normal fibroblast and MDA-MB-
231 human breast cancer cells inside the etched features. Fibroblast
cells stretch to avoid curved sidewalls of isotropic microchambers.
Cancer cells attach and spread on isotropic microchambers and deform
to take the shape of the curved sidewalls. These cells are less
discriminating on the substrate geometrical pattern. Scale bars
represent 10 μm
590 Biomed Microdevices (2009) 11:585–595
It is well known that cells sense the stiffness and the
geometry of their substrate, and modulate their cytoskeletal
organization and adhesion in response to the substrate
(Discher et al. 2005). Much of the previous work on
substrate stiffness has been conducted using 2-D routine
planar culture surfaces (Giannone and Sheetz 2006). With
respect to the substrate geometry, researchers have explored
the role of cytoskeleton on cell alignment using micro-
grooved substrates (Clark et al. 1991; Dalby et al. 2003;
DenBraber et al. 1996; Matsuzaka et al. 2000; Oakley and
Brunette 1993; Oakley et al. 1997; Rajnicek et al. 1997;
Walboomers et al. 1998, 1999a, b; Wojciakstothard et al.
Fig. 7 Confocal images show-
ing the distribution of the actin
and nuclei of the HS68 fibro-
blast cells attached to flat surfa-
ces in their normal state (A, B)
and after adding cytochalasin D
for 10 (C, D) and 30 min (E, F).
After 10 min, some stress fibers
were observed within the cyto-
skeleton of the cells (arrows).
After 30 min, the dense mesh
and long fibers of actin were
entirely disrupted through the
cell body. Scale bars represent
Biomed Microdevices (2009) 11:585–595 591
1995). Herein, we investigated the distinctive behavior of
HS68 normal fibroblast cells and MDA-MB-231 human
breast cancer cells on 3-D silicon isotropic environments
and investigated whether cytoskeletal structure (i.e. actin
filaments) influenced their adhesion.
Fibroblast cells adhered well to the silicon substrate due
to the abundance of parallel and cross-link formation of
actin filaments. These cells tended to maintain an organized
cytoskeleton and a state of high tension produced by
formation of stress fibers, lamellipodia, and focal adhe-
sions. This limited the ability of these cells to bend inside
the microchambers and adapt to the curved sidewalls.
Under these conditions, the whole cell body is a pre-
stressed tensegrity structure where the actin cytoskeleton
is in tension and the microtubules are in compression
(Ingber 1993). The high tension within the actin cytoskel-
eton is balanced by both the microtubules compression and
cell adhesion to the rigid silicon substrate. The balance
between cytoskeletal pre-stress and the mechanical force
within the cell body is the key determinant of cell shape
stability (Ingber 2003) (i.e., stretching and avoiding the
After 10 min treatment with cytochalasin D, some actin
stress fibers were still present in the cell cytoskeletal
structure. However, by 30 min, fibroblast cells lose their
entire actin filaments (i.e. stress fibers) and consequently
their cytoskeletal pre-stress. This primarily results in a
significant alteration in their morphology towards shape
instability. It is well known that the mechanical structure of
the cells, such as their elasticity and overall strength, is
dominated by actin filaments (Ananthakrishnan et al. 2006;
Stossel 1984; Tseng and Wirtz 2001). Specifically, there is
a direct linear correlation between the cell stiffness and the
amount of cellular pre-stress (Wang et al. 2002). Disruption
of the cell cytoskeleton causes a major reduction in cell
stiffness (Rotsch and Radmacher 2000). Therefore, after
adding cytochalasin D, these cells transform to a low
tensional state, their stiffness significantly decreases, and
therefore, they easily deform to adapt themselves with the
curved shape of 3-D microstructures—similar to the behavior
observed in breast cancer cells. Such adaptation is a sign of
their weak cytoskeleton after adding cytochalasin D. It is
worth mentioning that low tensional state of fibroblast cells
has been considered previously as the major cause for the
Fig. 8 Confocal images show-
ing the distribution of the actin
and nuclei of the MDA-MB-231
breast cancer cells in their nor-
mal state (A, B) and 10 min
after adding cytochalasin D (C,
D). Scale bars represent 20 μm.
In control condition, the actin
cytoskeleton of MDA cells were
distributed in a form of concen-
tric rings around the cell’s
nuclei. After 10 min treatment
with cytochalasin D the actin
cytoskeleton is completely dis-
rupted. Scale bars represent
592Biomed Microdevices (2009) 11:585–595
round/bipolar or dendritic morphology of these cells in 3-D
collagen matrices (Rhee and Grinnell 2007). Although not a
direct correlation between collagen matrices and our
isotropically etched silicon substrates, the round/dendritic
morphology of the fibroblasts after adding cytochalasin D in
our experiments comes from the same origin; i.e., cells’ low
tensional state by disrupting actin microfilaments.
In contrast to normal fibroblast cells, the absence of
stress fibers was evident in the cytoskeletal structure of
cancer cells. This suggests that the adhesion of the
cancerous cells to the silicon substrate is not as strong as
normal fibroblast cells. Previously it was also reported that
the presence of stress fibers is necessary for cell adhesion to
the substrate (Lewis et al. 1982; Soranno and Bell 1982).
Cancer cells adopted with round and spread shapes on flat
and curved sidewalls. Addition of cytochalasin D caused
these cells to become more rounded while it did not
significantly change their adhesion behavior to the 3-D
microstructures. This can be explained by the fact that even
in their original state, these cells, attached to silicon
substrates, did not possess actin stress fibers within their
cytoskeletal structure. One characteristics of cells after
oncogenic transformation is the loss of anchorage depen-
dent growth (Stoker et al. 1968). Transformed cells are less
responsive to surface rigidities (Wang et al. 2000), while
the normal cells probe substrate rigidity as a mechanical
feedback to modulate their shape and growth. Since
destabilization of the stress fibers and focal adhesion in
malignant cells enhances the migration process (Giannone
Fig. 10 Quantitative data showing the effect of the cytochalasin D on
fibroblast cell morphology 10 and 30 min after adding cytochalasin D.
The data show that after 30 min, there is a significant change in
fibroblast cell morphology from stretched in their normal state to
round (associated with dendritic extension)
Fig. 9 SEM images of the HS68 fibroblast and MDA-MB-231 cancer
cells after adding cytochalasin D. After 10 min of disrupting actin
cytoskeleton, fibroblast cells lost their anchorage dependent adhesion
to the etched microstructures. While after 30 min, the morphology of
these cells altered toward rounded morphology and their adhesion
became more like that of the cancer cells. Arrows show the dendritic
extension of the cells. Addition of cytochalasin D to cancer cells
caused the morphology of these cells to become more rounded while
there was not a significant alteration in their adhesion behavior to the
curved sidewalls. Scale bars represent 10 μm
Biomed Microdevices (2009) 11:585–595 593
and Sheetz 2006), it can be concluded that the observed
behavior of MDA-MB-231 cells on the fabricated 3-D
microstructures is a reflection of the metastatic nature of
these cells and the failure of their impaired cytoskeleton to
generate contractile forces on silicon substrates.
In summary, our research results support the hypothesis
that differential response of normal and cancer cells to the
various compartments of the 3-D silicon microenviron-
ments is a reflection of differences in their cytoskeletal pre-
stress and consequently their biomechanical properties.
Previously, the use of scanning force microscopy (SFM)
(Lekka et al. 1999) and optical tweezers (Guck et al. 2005)
provided quantitative measurements of the alterations in
mechanical properties of cancer cells compared to normal
cells. In general, these techniques showed that metastatic
cancer cells are significantly softer and less resistant to
deformation compared to normal cells, supporting our
hypothesis. Although such instrumentation provides valu-
able quantitative measurements of cellular mechanical
properties, we propose the precise surface topographical
patterns etched into silicon provide another approach to
evaluate biomechanical properties of different cell lines and
specifically predict their stiffness under normal and dis-
eased conditions. These microengineered structures can also
be used to assess dynamic cytoskeleton responses, and show
promise for applications in the identification of drugs able to
restore normal cytoskeleton function (Strobl et al. 2007).
Our future work will be focused on live measurement of
dynamical changes in cytoskeletal structure of normal and
cancer cells on the proposed microstructures.
In this work, we studied the behavior of individual HS68
normal human fibroblast and MDA-MB-2131 metastatic
human breast cancer cells under static cell culture con-
ditions in 3-D silicon microstructures. Our findings dem-
onstrated the contrasting behaviors of normal and cancer
cells in the fabricated 3-D microenvironments. Fibroblast
cells tended to stretch and maintained a more organized
cytoskeleton. The cancer cells were less discriminating in
response to the substrate geometry and deformed to take the
shape of the curved sidewalls of isotropic microchambers.
Treatment of the fibroblast cells with cytochalasin D,
significantly altered their morphology and adhesion inside
the etched features. However the adhesion of cancer cells to
the curved microstructures was not significantly altered
after disrupting their cytoskeleton using cytochalasin D.
Our results suggest that differences in the cytoskeleton
predominated by actin structures and biomechanical prop-
erties of these cells play a major role in their differential
adhesion to various compartments of the etched silicon
microstructures. The presented 3-D microstructures consti-
tute a versatile platform to evaluate biomechanical properties
of different cell types and to design novel cell-based assays
including anti-cancer drug testing. The utility of these
microstructures can be enhanced through the integration of
additional electrical and fluidic components.
Rajagopalan at the Department of Chemical Engineering at Virginia
Tech and the staff of Virginia Tech Micron Semiconductor Fabrication
Laboratory and Nanoscale Characterization and Fabrication Laboratory
(NCFL) and Institute for Critical Technology and Applied Sciences
The authors would like to thank Dr. Padma
R. Ananthakrishnan, J. Guck, F. Wottawah, S. Schinkinger, B.
Lincoln, M. Romeyke, T. Moon, J. Kas, J. Theor. Biol. 242,
502 (2006). doi:10.1016/j.jtbi.2006.03.021
C.J. Bettinger, B. Orrick, A. Misra, R. Langer, J.T. Borenstein,
Biomaterials 27, 2558 (2006). doi:10.1016/j.biomaterials.
J.L. Charest, A.J. Garcia, W.P. King, Biomaterials 28, 2202 (2007).
P. Clark, P. Connolly, A.S.G. Curtis, J.A.T. Dow, C.D.W. Wilkinson,
Development 99, 439 (1987)
P. Clark, P. Connolly, A.S.G. Curtis, J.A.T. Dow, C.D.W. Wilkinson,
Development 108, 635 (1990)
P. Clark, P. Connolly, A.S.G. Curtis, J.A.T. Dow, C.D.W. Wilkinson,
J. Cell Sci. 99, 73 (1991)
M.J. Dalby, M.O. Riehle, S.J. Yarwood, C.D.W. Wilkinson, A.S.G.
Curtis, Exp. Cell Res. 284, 274 (2003). doi:10.1016/S0014-4827
E.T. DenBraber, J.E. DeRuijter, H.T.J. Smits, L.A. Ginsel, A.F.
VonRecum, J.A. Jansen, Biomaterials 17, 1093 (1996). doi:10.
D.E. Discher, P. Janmey, Y.L. Wang, Science 310, 1139 (2005).
M.R. Dusseiller, D. Schlaepfer, M. Koch, R. Kroschewski, M. Textor,
Biomaterials 26, 5917 (2005). doi:10.1016/j.biomaterials.
K. Gantz, L. Renaghan, M. Agah, J. Micromech. Microeng. 18
G. Giannone, M.P. Sheetz, Trends Cell Biol. 16, 213 (2006).
J. Guck, S. Schinkinger, B. Lincoln, F. Wottawah, S. Ebert, M.
Romeyke, D. Lenz, H.M. Erickson, R. Ananthakrishnan, D.
Mitchell, J. Kas, S. Ulvick, C. Bilby, Biophys. J. 88, 3689
R.G. Harrison, J. Exp. Zool. 17, 521 (1914). doi:10.1002/
D.E. Ingber, J. Cell Sci. 104, 613 (1993)
D.E. Ingber, J. Cell Sci. 116, 1157 (2003). doi:10.1242/jcs.00359
R. Kalluri, M. Zeisberg, Nat. Rev. Cancer 6, 392 (2006). doi:10.1038/
L.D. Kojic, B. Joshi, P. Lajoie, P.U. Le, M.E. Cox, D.A. Turbin, S.M.
Wiseman, I.R. Nabi, J. Biol. Chem. 282, 29305 (2007).
M. Lekka, P. Laidler, D. Gil, J. Lekki, Z. Stachura, A.Z. Hrynkiewicz,
Eur. Biophys. J. Biophys. Lett. 28, 312 (1999)
L. Lewis, J.M. Verna, D. Levinstone, S. Sher, L. Marek, E. Bell. J.
Cell Sci. 53, 21 (1982)
594Biomed Microdevices (2009) 11:585–595
H.H. Liao, A.S. Andersson, D. Sutherland, S. Petronis, B. Kasemo, P.
Thomsen, Biomaterials 24, 649 (2003). doi:10.1016/S0142-9612
W.A. Loesberg, X.F. Walboomers, J.
Biomed. Mater. Res. Part A 75A, 723 (2005). doi:10.1002/jbm.
W.A. Loesberg, X.F. Walboomers, J.
Motil. Cytoskeleton 63, 384 (2006). doi:10.1002/cm.20132
J.Y. Mai, C. Sun, S. Li, X. Zhang, Biomed. Microdevices 9, 523
K. Matsuzaka, F. Walboomers, A. de Ruijter, J.A. Jansen, Clin. Oral
Implants Res. 11, 325 (2000). doi:10.1034/j.1600-0501.2000.
M. Nikkhah, J.S. Strobl, M. Agah, Biomed. Microdevices (2008).
C. Oakley, D.M. Brunette, J. Cell Sci. 106, 343 (1993)
C. Oakley, N.A.F. Jaeger, D.M. Brunette, Exp. Cell Res. 234, 413
A.M. Rajnicek, S. Britland, C.D. McCaig, J. Cell Sci. 110, 2905
S. Rhee, F. Grinnell, Adv. Drug Delivery Rev. 59, 1299 (2007)
C. Rotsch, M. Radmacher, Biophys. J. 78, 520 (2000)
Y.A. Rovensky, A.D. Bershadsky, E.I. Givargizov, L.N. Obolenskaya,
J.M. Vasiliev, Exp. Cell Res. 197, 107 (1991). doi:10.1016/0014-
C.E. Semino, J.R. Merok, G.G. Crane, G. Panagiotakos, S.G. Zhang,
Differentiation 71, 262 (2003). doi:10.1046/j.1432-0436.2003.
T. Soranno, E. Bell, J. Cell Biol. 95, 127 (1982). doi:10.1083/
M. Stoker, C. O’Neill, S. Berryman, V. Waxman, Int. J. Cancer 3, 683
T.P. Stossel, J. Cell Sci. 99, S15 (1984)
van Loon, J.A. Jansen, J.
van Loon, J.A. Jansen, Cell
J.S. Strobl, M. Nikkhah, M. Agah, in American Association for
Cancer Research Special Conference on Advances in Breast
Cancer Research: Genetics, Biology, and Clinical Applications,
(San Diego, CA, 2007).
T. Takezawa, Biomaterials 24, 2267 (2003). doi:10.1016/S0142-9612
R. Timpl, Curr. Opin. Cell Biol. 8, 618 (1996). doi:10.1016/S0955-
Y. Tseng, D. Wirtz, Biophys. J. 81, 1643 (2001)
A.M.P. Turner, N. Dowell, S.W.P. Turner, L. Kam, M. Isaacson, J.N.
Turner, H.G. Craighead, W. Shain, J. Biomed. Mater. Res. 51,
430 (2000). doi:10.1002/1097-4636(20000905)51:3<430::AID-
T.G. van Kooten, A.F.von Recum, Tissue Eng. 5, 223 (1999).
X.F. Walboomers, H.J.E. Croes, L.A. Ginsel, J.A. Jansen, Biomate-
rials 19, 1861 (1998). doi:10.1016/S0142-9612(98)00093-3
X.F. Walboomers, H.J.E. Croes, L.A. Ginsel, J.A. Jansen, J. Biomed.
Mater. Res. 47, 204 (1999a). doi:10.1002/(SICI)1097-4636
X.F. Walboomers, W. Monaghan, A.S.G. Curtis, J.A. Jansen, J.
Biomed. Mater. Res. 46, 212 (1999b). doi:10.1002/(SICI)1097-
H.B. Wang, M. Dembo, Y.L. Wang, Am. J. Physiol. Cell Physiol. 279,
N. Wang, I.M. Tolic-Norrelykke, J.X. Chen, S.M. Mijailovich, J.P.
Butler, J.J. Fredberg, D. Stamenovic, Am. J. Physiol. Cell
Physiol. 282, C606 (2002)
R.A. Weinberg, The Biology of Cancer (Garland Science, Taylor &
Francis Group, New York, 2007)
B. Wojciakstothard, A.S.G. Curtis, W. Monaghan, M. McGrath, I.
Sommer, C.D.W. Wilkinson, Cell. Motil. Cytoskelet. 31, 147
Biomed Microdevices (2009) 11:585–595 595