Available via license: CC BY 4.0
Content may be subject to copyright.
J. Funct. Biomater. 2015, 6, 241-258; doi:10.3390/jfb6020241
Journal of
Functional
Biomaterials
ISSN 2079-4983
www.mdpi.com/journal/jfb
Article
Differential Cell Adhesion of Breast Cancer Stem Cells on
Biomaterial Substrate with Nanotopographical Cues
Kenneth K.B. Tan
1,†
, Christine S.Y. Giam
2
, Ming Yi Leow
2
, Ching Wan Chan
3,
* and
Evelyn K.F. Yim
1,2,3,†,
*
1
Mechanobiology Institute, National University of Singapore, T-Lab, #05-01, 5A Engineering Drive 1,
Singapore 117411; E-Mail: mbiktkb@nus.edu.sg
2
Department of Biomedical Engineering, National University of Singapore, EA-03-12, 9 Engineering
Drive 1, Singapore 117575; E-Mails: zest228@gmail.com (C.S.Y.G.);
leowmingyi@gmail.com (M.Y.L.)
3
Department of Surgery, Yong Loo Lin School of Medicine, National University of Singapore,
NUHS Tower Block, Level 8, 1E Kent Ridge Road, Singapore 119228
†
These authors contributed equally to this work.
* Authors to whom correspondence should be addressed; E-Mails: ching1chan@gmail.com (C.W.C.);
eyim@nus.edu.sg (E.K.F.Y.); Tel.: +65-6516-7322 (E.K.F.Y.); Fax: +65-6872-3069 (E.K.F.Y.).
Academic Editor: Akhilesh K. Gaharwar
Received: 12 February 2015 / Accepted: 15 April 2015 / Published: 21 April 2015
Abstract: Cancer stem cells are speculated to have the capability of self-renewal and
re-establishment of tumor heterogeneity, possibly involved in the potential relapse of cancer.
CD44
+
CD24
−/low
ESA
+
cells have been reported to possess tumorigenic properties, and these
biomarkers are thought to be highly expressed in breast cancer stem cells. Cell behavior can
be influenced by biomolecular and topographical cues in the natural microenvironment.
We hypothesized that different cell populations in breast cancer tissue exhibit different
adhesion characteristics on substrates with nanotopography. Adhesion characterizations
were performed using human mammary epithelial cells (HMEC), breast cancer cell line
MCF7 and primary invasive ductal carcinoma (IDC) cells obtained from patients’ samples,
on micro- and nano-patterned poly-L-lactic acid (PLLA) films. Topography demonstrated a
significant effect on cell adhesion, and the effect was cell type dependent. Cells showed
OPEN ACCESS
J. Funct. Biomater. 2015, 6 242
elongation morphology on gratings. The CD44
+
CD24
−/low
ESA
+
subpopulation in MCF7 and
IDC cells showed preferential adhesion on 350-nm gratings. Flow cytometry analysis
showed that 350-nm gratings captured a significantly higher percentage of CD44
+
CD24
−
in
MCF7. A slightly higher percentage of CD44
+
CD24
−/low
ESA
+
was captured on the 350-nm
gratings, although no significant difference was observed in the CD44
+
CD24
−
ESA
+
in IDC
cells across patterns. Taken together, the study demonstrated that the cancer stem cell
subpopulation could be enriched using different nanopatterns. The enriched population could
subsequently aid in the isolation and characterization of cancer stem cells.
Keywords: breast cancer; cancer stem cells; nanotopography; label-free isolation; cell adhesion
1. Introduction
Despite recent improvements in breast cancer mortality, many patients relapse after initial response
to conventional therapy and chemotherapy. Several alternative hypotheses have been proposed to
explain this treatment failure and recurrence. Of particular interest is the cancer stem cell theory, which
suggests that a small subpopulation of cells within tumors possess the unique ability to self-renew and
generate the diverse cells that comprise the tumor [1]. Often referred to as cancer stem cells (CSCs),
these cells are found to be resistant to therapy and, hence, may reinitiate tumor growth after
treatment [1–3]. As such, isolation of this tumorigenic population from tumor tissues will be crucial for
therapy development. Breast CSCs were first identified and isolated by Al-Hajj et al. [4]. When human
breast tumors were propagated in non-obese diabetic severe combined immunodeficiency (NOD/SCID)
mice, it was observed that as few as 100 of the CD44
+
CD24
−/low
cells were able to give rise to new
tumors. Thus, breast CSCs are postulated to have a CD44
+
CD24
−/low
phenotype, strongly expressing the
adhesion CD44 molecule, while very weakly expressing the adhesion CD24 molecule. Subsequent
studies were conducted to validate Al-Hajj’s findings [5–7].
The CSC fractions in solid tumors have been observed to be highly impure, and thus, the reported
frequencies for the same tumor types have varied enormously between different research groups [1].
Thus, more definitive markers are required to better characterize the CSCs. In the breast CSC field, it
was observed from Abraham’s work that the proportion of CD44
+
CD24
−/low
cells ranged from 0 to 40%
in normal tissues [5]. Human breast cancer cell lines also differ quantitatively in the proportion of
CD44
+
CD24
−
cells [7]. The percentage of CD44
+
CD24
−
cells within breast cancer cell lines was found
to be uncorrelated with tumorigenicity [8]. In addition, studies have found that the epithelial specific
antigen (ESA) cell surface molecule is overexpressed by the majority of human epithelial carcinomas,
including breast carcinomas [8–10]. The tumorigenic activity of the CD44
+
CD24
−/low
population was further
enhanced when the CD44
+
CD24
−/low
ESA
+
cell population was used [4]. Therefore, CD44
+
CD24
−/low
ESA
+
is
a potential breast CSC marker for further investigation of this group of breast cancer cells with
tumorigenic properties. The methods used to isolate CSCs have generally revolved around the use of
labels, such as selective expression of surface markers [1,4,11]. However, while this method enables the
identification and isolation of a specific sub-group of CSC, the marker may not be able to serve as a
universal marker for CSC.
J. Funct. Biomater. 2015, 6 243
A second method of CSC isolation is the Hoechst dye exclusion. These cells are referred to as side
population cells and have been found to exhibit CSC characteristics of tumorigenicity and chemotherapeutic
drug resistance [12,13]. A third method based on the observation that CSCs lack 26S proteasome
function has emerged [3,14]. Cells with reduced 26S proteasome activity express CSC markers and
appear to be more tumorigenic than the control cells. They have also been demonstrated to characterize
a sub-population of the CD44
+
CD24
−/low
breast CSCs [15].
Label-free methods exploiting the differences in physical properties, such as cell size, density, cell
adhesion and dielectric properties, are also being explored. Advantages conferred by label-free methods
include less laborious and time-consuming procedures, as the preparation for staining before and after
cell separation will not be needed. Nonetheless, one potential drawback of these methods is that these
physical differences might be insufficient for accurate cell separation. Substrates act as intelligent
surfaces capable of providing biochemical and topographical signals to guide cell adhesion, spreading,
morphology, proliferation, and eventually, cell differentiation [16]. Numerous studies have demonstrated
that cell adhesion behavior is significantly affected by surface nanotopography [17,18]. Cell adhesion
can be increased or decreased by changing the material or geometry used to construct the surface
structure [19]. Kwon and co-workers [19] used a nanotopographic substrate in a microfluidic approach
to separate human breast cancer cells using cell adhesion as a physical marker. The breast cancer cell
line (MCF7) and normal human breast epithelial cells (MCF10A) were observed to display different
adhesion properties when cultured on substrates with different topographies. However, the cell adhesion
analysis of the cell line mono-culture may differ from the heterogenicity of breast cancer tissues, and
therefore, the cell adhesion properties of a heterogeneous population would need to be further investigated.
Thus, we hypothesize that different cell populations in the heterogeneous breast cancer tissue will
respond differently to nanotopographical cues, potentially providing a means of enriching breast CSCs
for isolation. The information can be used to develop a fast and easy method to isolate cancer stem cells,
as well as different cell populations from cancer tissue, via a label-free method employing the differences
in cell adhesion. Here, we investigated the adhesion characteristics of human mammary epithelial cells
(HMEC), invasive breast cancer cell line MCF7 and invasive ductal carcinoma (IDC) primary breast
tumor cells on nanopatterned poly-L-lactic acid (PLLA) films, as well as studying the effectiveness of
these nanopatterns in preferential adhesion and isolation of breast CSCs.
2. Results
2.1. Fabrication of Patterned Poly-L-Lactic Acid Films
Solvent casting was used to make the PLLA (3% w/v in chloroform) films from the
poly(dimethylsiloxanes) (PDMS) masters with topographies of 250-nm wells, 1-µm wells, 350-nm
gratings, 1-µm gratings and 1-µm pillars. PLLA solution was cast on the PDMS master. After drying
overnight at room temperature and a brief vacuum drying to remove any residual solvent, the PLLA film
was gently stripped off. The scanning electron micrographs (SEM) of the patterned PLLA films showed
that the fabrication techniques used were able to replicate nanostructures accurately (Figure 1), verifying
the fidelity of the replication.
J. Funct. Biomater. 2015, 6 244
Figure 1. Scanning electron microscopic (SEM) images of poly-L-lactic acid (PLLA)
patterned films with nano- and micro-structures.
2.2. Influence of Surface Topography on Cell Morphology
Immunofluorescence staining was used to examine the cell morphology and to assess the expression
of CD44, CD24 and ESA qualitatively. MCF7 and HMEC have different cell morphologies when seeded
on substrates with nanotopography after 24 h of culture (Figure 2). Both cell types attached well and in
clusters onto the patterned PLLA films without pre-coating with extracellular matrix protein. While a
round morphology was most commonly observed, MCF7 and HMEC were well spread on patterned
PLLA compared to unpatterned PLLA (Supplementary Figure S1), and cells were slightly more spread
on 1-µm wells, 250-nm wells and 1-µm gratings, although a significant difference was not observed
among patterns. Cell elongation was observed on the grating patterns, showing that topography was
interacting with cells and might affect differential adhesion within heterogeneous cell populations.
Figure 2. Immunofluorescence images of (A) MCF7 and (B) human mammary epithelial
cells (HMEC) after 24 h of culture. The cells were stained for CD44 (green) and co-stained
with either CD24 (red) or ESA (red), with DAPI as the counter-stain of nuclei (blue).
J. Funct. Biomater. 2015, 6 245
In the heterogeneous population of MCF7 after 4 h of culture, the degree of surface marker expression
varied across the population, but generally, the majority of HMECs were observed to be CD44
+
, CD24
+
and ESA
+
across all of the patterns. On the other hand, the expression of CD44, CD24 and ESA in MCF7
cells varied between different patterned samples (Supplementary Figure S2). The heterogeneity in the
immunophenotype of the MCF7 cells after a 24-hour culture period was also clearly reflected in Figure 2.
The differences in the subpopulations were further analyzed with cell counting and flow cytometry.
2.3. Differential Adhesion of MCF7 and HMEC on Patterned PLLA Films
The number of attached cells after a certain time point is a key parameter in determining the adhesion
behavior of the cells. Cell adhesion is drastically influenced by the surface topography, thus affecting
the number of cells that adhere to the substrate [20,21]. The number of attached cells on each of the
PLLA films at the time points of 4 and 24 h is presented in Figures 3A and 4A.
Figure 3. Effect of nanotopography on MCF7 cell adhesion 4 h after seeding on PLLA films.
(A) Number of MCF7 attached on PLLA films, seeding density 16,000 cells/cm
2
. The 1-μm
well has more attached cells, and the 350-nm grating has less attached cells
(***P < 0.001,
mean ± SD, n = 5); (B) MCF7 culture after 4 h showed that the 350-nm grating captured a
significantly higher percentage of CD44
+
CD24
−/low
ESA
+
cells (*P < 0.05, **P < 0.01,
mean ± SD, n = 3); (C) Flow cytometry data showed that the 350-nm grating captured a
higher percentage of CD44
+
CD24
−
cells. The percentage of CD44
+
CD24
−
cell population in
MCF7 cells 4 h after seeding on PLLA films (*P < 0.05, mean ± SEM, n = 3).
Two different experimental setups were used to measure the number of adhered MCF7 cells on the
PLLA film. In the first one, to assess the total number of adhered cells, cells were seeded on PLLA film;
after the experimental time points (4 and 24 h), cells were detached enzymatically using trypsin/EDTA
and counted using a hemocytometer. In another experimental setup to assess the percentage of
CD44
+
CD24
−/low
ESA
+
in the population, the cells and nuclei were fluorescently stained. Images were
taken of random areas (at least five per sample) and counted.
J. Funct. Biomater. 2015, 6 246
Figure 4. (A) Number of MCF7 attached 24 h after seeding on PLLA films, seeding density
16,000 cells/cm
2
(*P < 0.05, **P < 0.01, mean ± SD, n = 5); (B) Number of HMEC attached
24 h after seeding on PLLA films, seeding density 16,000 cells/cm
2
(*P < 0.05, mean ± SD,
n = 5). Percentage of CD44
+
(C) MCF7 and (D) HMECs from HMEC-MCF7 co-culture.
The graph represents the percentage of CD44
+
cells (in the CD44
+
/CellTracker Red
−
region
for MCF7 or the CellTracker Red
+
region for HMEC) in an event count of at least 5000 after
the 24-hour culture period on the PLLA films. Co-cultures of HMEC and MCF7 were seeded
on PLLA films, and after 24 h of culture time, cells were fixed and stained for CD44.
After 4 h of culture of MCF7 cells on the various PLLA films, the 1-μm well had more attached cells
compared to all samples and the control (unpatterned) (P < 0.001, mean ± SD, n = 5) (Figure 3A). Cell
counting after 4 h of culture for CD44
+
CD24
−/low
ESA
+
MCF7 cells showed that a higher number of
CD44
+
CD24
−/low
ESA
+
MCF7 cells were captured on the nanoscale gratings, as compared to the control
and other micro-patterns (P < 0.05, n = 3) (Figure 3B). Flow cytometry analysis also showed that
350-nm gratings captured a significantly higher percentage of CD44
+
CD24
−
than the unpatterned control
J. Funct. Biomater. 2015, 6 247
(P < 0.05, n = 3). There was no significant difference in the percentage of CD44
+
CD24
−
cell population
in MCF7 across different patterns (Figure 3C).
After 24 h of culture on the various PLLA films, MCF7 cell counting showed that 1-µm wells and
250-nm wells had more attached cells than other samples (Figure 4A). The 1-μm wells had significantly
more cells than the 350-nm gratings (P < 0.01, mean ± SD, n = 5), and similarly, the 250-nm wells had
more cells than the 350-nm grating (P < 0.05, mean ± SD, n = 5). Comparing between the gratings, the
1-µm pattern has more attached cells than the 350-nm grating pattern, but the difference was not
statistically significant.
As HMECs were cultured in serum-free medium, which might delay cell attachment, only a low
number of cells had attached after 4 h. Even when using a more sensitive method in counting the
DAPI-stained nuclei, low numbers of HMECs were attached on the patterned samples (Supplementary
Figure S3), and a large variation was observed among replicas. Thus, the HMEC number was assessed
24 h after seeding with the enzymatic detachment method. The HMEC number after 24 h of culture on
the various PLLA films (Figure 4B) showed that the 1-μm wells had fewer attached HMECs as compared
to the 1-μm grating (P < 0.05, mean ± SD, n = 5). The highest number of HMEC adhered on the 1-μm
gratings, while the lowest was noted in the 1-µm wells at the 24-hour time point.
To better mimic a heterogeneous mixture of mammary epithelial cells and cancer stem cell population
in tumors, HMECs and MCF7 were co-cultured. Flow cytometry was performed on the co-culture to
further investigate the adhesion properties of the two cell types with the objective of identifying the
preferential adhesion of the breast cancer stem cells on different patterned PLLA. In the interpretation
of the data, the HMEC was stained with CellTracker Red. The population of CellTracker Red-negative
(MCF7) and CD44
+
(mean ± SD, n = 3) in each sample was represented in the Figure 4C. The 350-nm
gratings had a significantly higher percentage of CD44+ cells compared to the control (P < 0.05). For
the 1-μm well, the percentage of CD44
+
cells ranged from 0.06% to 26.4% with a standard deviation of
14.8%, which indicates a higher degree of inaccuracy or inconsistency in the readings. Meanwhile, the
population of CellTracker Red-positive (HMEC) and CD44
+
(mean ± SD, n = 3, Figure 4D) also showed
that the 350-nm gratings and 1-µm gratings captured slightly higher percentages of CD44
+
than the other
samples, but no significant difference was observed among the samples.
2.4. Characterization of IDC Cells Adhered on Patterned PLLA Films
Invasive ductal carcinoma (IDC) cells, Passage 3–5, from breast tumor tissues of patient donors were
obtained to investigate the preferential cell adhesion on different patterned PLLA films. To characterize
the cells seeded on different patterns, cell counting, immunofluorescence staining and flow cytometry
were employed to analyze the cell morphology and surface markers.
IDC cells attached well on the PLLA films 4 h after cell seeding in serum-free media. The IDC cells
seeded on gratings were widely dispersed and had pronounced elongation and alignment in the direction
of the gratings (Figure 5). The morphology of the primary cells (Figure 5) and the MCF7 cells
(Figure 2) was very diverse. MCF7 cells cultured on the patterned films were flat with a spherical
nucleus. By contrast, the primary cells showed stellate morphology, especially on the 250-nm wells and
to a lesser extent on the 350-nm gratings.
J. Funct. Biomater. 2015, 6 248
Figure 5. Immunofluorescence images of primary invasive ductal carcinoma (IDC) cells on
(A) unpatterned PLLA film; (B) 250-nm wells and (C) 350-nm gratings after 4 h of culture.
Blue: DAPI; green: CD44; red: CD24; magenta: ESA. Original magnification: 20×.
Bar = 100 μm.
As shown in the MCF7 immunostaining (Figure 2), heterogeneity was also observed in the
immunophenotype of the primary cells (Figure 5) on the various patterned films after 4 h of culture.
Surface marker expression varied across the population. Generally, most of the primary cells were
observed to be CD44
+
, CD24
+
and ESA
+
. Only minute numbers of them were CD44
+
CD24
−/low
ESA
+
.
The preferential adhesion behavior of the CD44
+
CD24
−/low
ESA
+
primary IDC cells on the nanoscale
features was evident from the fluorescence images.
Due to the variations among patient samples, primary IDC cell counts varied significantly from one
batch to another. The numbers of attached IDC cells on patterned samples were normalized to the number
of IDC cells attached on the unpatterned control to facilitate the comparison over three different patient
samples. Furthermore, due to the limited cell numbers available from each patient, the cell counting of
the total cell number was performed using image analysis of the CellTracker Red- and DAPI-stained
cells. From the cell counting, the 250-nm wells showed a significantly higher number of attached primary
IDC cells, compared to the 350-nm gratings (P < 0.001) and unpatterned control (Figure 6A, P < 0.01).
No significant difference in the normalized cell numbers was observed among the other patterns.
Because both nanoscale patterns (350-nm gratings and 250-nm wells) captured a higher number of
CD44
+
CD24
−/low
ESA
+
cells from the MCF7 mono-culture at the 4-hour time point (Figure 3B), these
two nano-patterns were chosen for cell counting of the immuno-stained population in the IDC cell
experiment. The results showed that a significantly higher number of CD44
+
CD24
−/low
ESA
+
cells were
captured on the 250-nm well (Figure 6B). From the immunostaining observation, the number of
CD44
+
CD24
−/low
ESA
+
cells increased with the total cell population in the 250-nm wells; however, in
terms of percentage, the CD44
+
CD24
−/low
ESA
+
expression population was higher on gratings
(Figure 6C). The 35-0nm gratings captured a significantly higher percentage of CD44
+
CD24
−/low
ESA
+
compared to unpatterned control (P < 0.05).
Flow cytometry was conducted twice on two batches of the IDC cells (Figure 6D). ANOVA
(Kruskal–Wallis test followed by Dunn’s post hoc test) results did not show any significant difference among
different patterned samples. However, a slightly higher percentage of CD44
+
CD24
−/low
ESA
+
was observed
on the 350-nm gratings, which agrees with our observations from counting immunofluorescently-stained
samples (Figure 6C).
J. Funct. Biomater. 2015, 6 249
Figure 6. (A) Normalized number of IDC cells attached 4 h after seeding on PLLA films
(**P < 0.01, ***P < 0.001, mean ± SD, n = 3; except for 1-µm pillars, where n = 1). The
IDC cell adhesion behavior normalized to unpatterned control shows preference for
nanoscale topography with the 250-nm wells. (B) The normalized number of primary IDC
cells cultured after 4 h showed the 250-nm wells as the best topography to capture a
significantly higher numbers of CD44
+
CD24
−/low
ESA
+
cells (***P < 0.001, mean ± SD,
n = 3). (C) By immunofluorescence staining and cell counting, the percentage of
CD44
+
CD24
−/low
ESA
+
showed that
350-nm gratings captured a significantly higher
percentage of CD44
+
CD24
−/low
ESA
+
cells (*P < 0.05, mean ± SD, n = 3). (D) Flow
cytometry analysis of primary IDC cell culture after 4 h showed that 350-nm gratings
captured a higher percentage of CD44
+
CD24
−/low
ESA
+
cells, but the difference was not
significant (mean ± SD, n = 2; except for 1-µm pillars, where n = 1).
Comparing the total number of adhered cells and the CD44
+
cells captured at 4-hour and 24-hour time
points, the trends of the preferential adhesion of MCF7 were comparable at both time points. The
different results between the IDC cell adhesion and MCF7 cell adhesion might be due to several factors.
First, as the fluorescent staining required more washing steps, cells could have been washed away due
to the varied strength of cell attachment to PLLA. Furthermore, the seeding densities used were different.
Despite the discrepancy between both setups, we showed that MCF7, HMEC, IDC cells and the
CD44
+
CD24
−/low
ESA
+
cells among the cell populations have preferential attachment on different patterns.
3. Discussion
Cancer stem cells (CSC) are a population of proliferating cells with the capability of self-renewal and
re-establishment of tumor heterogeneity [22]. Since the establishment of the cancer stem cell theory,
increasing evidence indicates that cancer stem cells play an important role in tumorigenesis and potential
relapse of cancer. However, in general, current cancer therapies, such as chemotherapy, target rapidly
proliferating cells instead of CSCs. As such, many have suggested that in order to reduce cancer relapse,
eradication of CSCs is crucial.
As cells are surrounded by submicron features in their microenvironment, it has been shown that
nanotopography influences cell adhesion, proliferation and differentiation [23]. We hypothesize that
J. Funct. Biomater. 2015, 6 250
different cell populations in breast cancer tissue exhibit different adhesion characteristics on substrates
with nanotopography and investigated the adhesion characteristics of three cell populations, HMEC,
breast cancer cell line MCF7 and primary IDC cells, on biomaterial substrates with nanotopographical
cues. Based on the adhesion difference, the cancer stem cell-like population in MCF7, HMEC and IDC
cells was separated and enriched from the mono-culture or the co-culture of HMEC and MCF7 using
nanotopography and CSC surface markers. The results have shown a trend of preferential cell adhesion
in both MCF7 and primary IDC cells cultured on different nanotopography after 4 and 24 h. Among the
topographies tested, the nanoscale features of 350-nm gratings are consistently the most effective in
enriching the CD44
+
CD24
−/low
ESA
+
breast CSCs. Interestingly, the total cell population has preferential
adhesion to 1-μm gratings and 250-nm wells for MCF7 and primary IDC cells, respectively, suggesting
that sub-populations of cancer cells exhibit different adhesion characterization to topographical structures.
Dick and colleagues’ studies in leukemic cells in the early 1990s gave one of the first descriptions of
cancer stem cells [19]. They were able to isolate a subpopulation of leukemic cells via biomarkers
CD34
+
/CD38
−
and initiated acute myeloid leukemia in NOD/SCID mice using these leukemic
cells [22,24]. CSCs were also found in the solid tumors of breast cancer [4,22,25,26]. Breast cancer cells
have demonstrated not only self-renewal capability, but also a wide spectrum of progeny [25]. The
ESA
+
CD44
+
CD24
−/low
lineage cells isolated from human breast tumors were able to form tumors in
eight-week-old NOD/SCID mice [4]. These cells formed a very small proportion of the total cell
population, and as few as 200 cells were able to consistently form tumors in the mice and, thus, were
identified as breast CSCs [4,25].
Although researchers were able to isolate these putative CSCs, a principle to isolate stem cells from
any kind of cancer has yet to be established. In the case of breast cancer, there remains a need to
determine whether CD44
+
CD24
−
ESA
+
cells are true breast CSCs across various breast cancer subtypes.
In addition, reports have confirmed that in both breast cancer-derived cell lines and breast tumors,
CD44
+
CD24
−
phenotypes were not necessarily associated with patient outcome or the ability to
metastasize [7,27]. Therefore, further research is required to identify and separate CSCs in breast cancer.
Meanwhile, current therapies fail to account for potential differences in drug sensitivity or target
expression between the CSCs and the more frequent non-tumorigenic cells [28]. The ability to isolate
breast CSCs enables the testing of the effectiveness of cancer drugs that are currently used and will also
enable the development of therapeutic strategies aimed at selectively targeting breast CSCs [22].
Current methods for detecting and separating CSCs from solid tumors have been based on the
observation of morphological features or labeling with specific markers [25]. The labeling-assisted
method is more commonly used and is highly specific. However, it requires laborious procedures prior
to separation and time-consuming post-processing, such as the removal of labels [29]. Label-free
methods have been studied to separate cancer cells, which utilize differences in physical properties, such
as cell size, density [20], cell adhesion [20,30] and dielectric properties [20], but the physical distribution
is not high enough for efficient and accurate separation of cell populations, which limits their use. Kwon
and co-workers used nanotopographic substrates to separate and enrich human breast cancer cells [19].
They demonstrated that normal human breast epithelial cells (MCF10A) have stronger adhesion to the
substratum than MCF7, regardless of culture time, surface topography and flow rates. This separation
technique was reported to have 78% efficiency [19].
J. Funct. Biomater. 2015, 6 251
Kwon’s study also found that the separation of human breast cancer cells from a mixed phase of
MCF7 and MCF10A cells was optimized to occur after 2 h of culture on 400-nm gratings.
A shorter duration of 2 h was chosen, as the high flow rate can exert significant shear damage to the
adhered cells under longer culture periods. In contrast, in our static cultures, 4-hour is a reasonable
duration, as the cells might not be able to recognize the surface within a short period of time. The
4-hour time point also appears to be the optimum duration for the separation and enrichment of the breast
CSCs (as denoted by the phenotype of CD44
+
CD24
−/low
ESA
+
) from the heterogeneous MCF7 cancer
cells and the primary cells under static culture conditions. This is in line with the development of a
label-free method to isolate CSCs, as a shorter culture and analysis time is more desirable for cell analysis
and cell isolation.
Kwon and colleagues determined that the use of 400-nm gratings resulted in the generation of the
largest shear stress on the cell surface, facilitating the detachment of the MCF7 cells. Thus, it is the
preferential detachment behavior that allowed the successful separation of the breast cancer cells in
Kwon’s study [19]. On the contrary, we have focused on the preferential attachment behavior of the
cells, which may account for the difference of the choice of topographical feature. Thus, future work
may involve characterizing the unattached cells, as well, to establish the most effective nanotopography
for the isolation of the CSCs.
The mixture of MCF7 and MCF10A used by Kwon served to mimic the primary tissue of a breast
cancer tumor; it has been proven that although breast and breast cancer cell lines have adapted to tissue
culture conditions, they retain many of their phenotypic and genotypic properties over many passages.
Hence, despite their acquired ability to grow in vitro, cell lines continue to share many of the molecular
and genetic features of the primary breast cancers from which they were derived [8]. This validates the
use of MCF7 to elucidate the preferential adhesion behavior of tumorigenic breast CSCs.
In this research, cell adhesion is evaluated based on the number of cells that adhered to the substrate
and the cell morphology. From the data presented, less MCF7 and IDC cells adhered to gratings as
compared to wells. From immunostaining, the MCF7 on the grating displayed higher fluorescence
intensity for CD44 staining and a higher CD44
+
CD24
−/low
ESA
+
population, as compared to wells and
the control. By counting the number of CD44
+
cells that adhered to the substrates, more than 40% of the
cells that the 350-nm gratings captured were CD44
+
CD24
−/low
ESA
+
MCF7 cells, suggesting that the
350-nm grating is most effective in the enrichment of CD44
+
CD24
−/low
ESA
+
MCF7 in a monoculture.
The difference in cell adhesion is a key parameter to enable isolation of breast CSCs from the
heterogeneous cell suspension. Cell adhesion is affected by surface chemistry and the topography of the
polymer substrate, as well as depending on the geometric parameters of the nanotopographic substrate.
To mimic the heterogeneous cell population of a breast cancer tumor, the differential adhesion behavior
of the co-culture of MCF7 and HMEC was studied. The nanotopography that would be the most suitable
for the intended purpose would be one that captures the most CD44
+
cells in the heterogeneous
co-culture experiment. The MCF7 that adhered on 350-nm gratings continue to give high fluorescence
intensity when stained for CD44, as compared to the well pattern, and have the highest percentage of
cells that are CD44
+
when analyzed by flow cytometry. This most likely indicates that in the co-culture,
the nanotopography continues to affect the CD44
+
population in MCF7 and HMEC by the same
mechanism as when MCF7 was cultured alone, thus displaying a similar behavior.
J. Funct. Biomater. 2015, 6 252
Although the three different cell populations, MCF7, HMEC and IDC, showed different preferences
in terms of the total cell adhesion on the 1-µm well, 1-µm gratings and 250-nm wells, respectively, the
highest proportion of CD44
+
CD24
−/low
ESA
+
was consistently captured on the 350-nm gratings. The
observations supported our hypothesis that the different cell populations in the heterogeneous breast
cancer tissue showed different preferential adhesion to various topographical patterns. In particular, the
CD44
+
CD24
−/low
ESA
+
CSC population showed a preferential retention and enrichment on 350-nm
gratings, suggesting that this tumorigenic CSC population could be enriched by using nanoscaled
gratings. However, it is uncertain whether the 350-nm grating induces the expression of CD44 or if it is
able to isolate the cancer stem cell-like population in MCF7. This could be verified by studying whether
the side population isolated by 350-nm gratings is indeed CSCs.
Cell adhesion is directly affected by cell-extracellular substrate interaction; however, the mechanism
of the topographical regulation of cell adhesion remains unclear. One of the suggested mechanisms,
known as “self-induced mechanotransduction” by Dalby et al., is based on the hypothesis that surface
topography alters nuclear morphology and chromosome position in adherent cells, leading to changes in
gene transcription [31,32]. Another mechanism suggested by Ingber et al. is the “direct mechanotransduction”
mechanism, whereby forces encountered by cells during cell adhesion are directly transmitted to nucleus
via cytoskeleton, and altered cytoskeletal tension then feeds back to induce local changes in focal
adhesion assembly [33,34]. Our group has demonstrated that topography regulates cell morphology by
modulating the physical force-sensing of the cells [35,36]. Moreover, the focal adhesion signaling and
cytoskeletal contractility are essential for nanotopographical regulation in stem cell differentiation [35,37].
Given the short time points (4 and 24 h) used in this research, it is unlikely that cell behavior would
be affected by gene transcription. Preliminary data from cell adhesion counting and immunostaining
support the concept that nanotopography affects cell behavior via a direct mechanism. Moreover, the
morphology difference between the attached MCF7, HMEC and primary cells might provide a
mechanistic explanation of the preferential cell adhesion. MCF7, HMEC and IDC cells showed
elongated morphology on gratings patterns (Figures 2 and 5); however, MCF7 and HMECs showed
more pronounced cell spreading and elongation on gratings. Immunofluorescence staining of vinculin
(Supplementary Figure S4) illustrated the focal adhesion of the MCF7 and HMECs on the patterns at
24 h. MCF7 on micro-sized patterns showed fewer focal adhesions, as compared to the nano-sized
patterns, while HMECs showed well-established focal adhesion on all patterns. Further studies in cell
morphology with SEM could be conducted to examine cell-topography interactions to provide a better
understanding of the observed morphology differences. This will also illuminate the mechanism by
which nanotopography influences cell adhesion. The cell morphology of the CD44
+
CD24
−/low
ESA
+
cell
should be further analyzed to provide a better basis for comparison of the cancer stem cell preferential
adhesion among the different topographies.
In conclusion, between the two time points of four and 24 h, 4 h appear to be the optimum time mark
for isolating the CD44
+
CD24
−/low
ESA
+
tumorigenic cell population from both the MCF7 and the primary
cell cultures. In the MCF7 and IDC and MCF-7-HMEC co-culture cell experiments, the trend of
preferential adhesion behavior of these CD44
+
CD24
−/low
ESA
+
cells on the nanoscale features was
observed. Among the topographies tested, the 350-nm gratings are consistently the most efficient in
enriching the CD44
+
CD24
−/low
ESA
+
or breast CSCs from the MCF7 and primary IDC cell populations.
Moving forward, to develop a biomaterial substrate for effective cancer stem cell isolation, the isolated
J. Funct. Biomater. 2015, 6 253
cells can be further characterized and their cancer stem cell identity further studied, for instance, utilizing
the ALDH1 enzyme as a potential marker for CSCs. Scaling up of the isolation and expansion of the
isolated cells can also be further investigated.
4. Conclusions
We have successfully demonstrated the preferential adhesion of different breast cancer cell types on
micro- and nano-sized topographies. From the adhesion test results, preferential cell adhesion of MCF7
and primary IDC cells cultured on different nanotopographies for 4 and 24 h was observed. The MCF7
and IDC cells are sensitive to the dimensions of the patterns as judged by a significant difference in total
cell adhesion and the population of CD44
+
CD24
−/low
ESA
+
captured. Among the topographies tested, the
350-nm grating is consistently most effective at inducing the attachment of a larger percentage of MCF7
and IDC cells with our proposed CSC phenotype CD44
+
CD24
−/low
ESA
+
. Flow cytometry analysis using
two markers CD44
+
CD24
−/low
also demonstrated that the 350-nm grating captured a larger number of
CD44
+
CD24
−/low
cells. For primary IDC cells, immunofluorescence analysis and flow cytometry analysis
showed that the CD44
+
CD24
−/low
ESA
+
cells had greater adherence on the 350-nm gratings.
This study demonstrated that the subpopulations of cancer cells could exhibit different preferential
adhesion behaviors. Therefore, by carefully selecting the optimal topography, the sub-group of cancer
cells and CSCs could be separated by different topographies.
5. Materials and Methods
5.1. Fabrication of Micro- and Nano-Patterned Poly-L-Lactic Acid Films
A poly(dimethylsiloxane) (PDMS) soft mold was made from a micro- or nano-patterned template by
soft lithography, as previously described [38]. Briefly, the PDMS 10:1 elastomer and curing agent
mixture was degassed, dispensed onto the masters and cured at 65 °C.
PLLA films were made by solvent casting using 3% w/v PLLA (Sigma Aldrich, St. Louis, MO, USA)
dissolved in chloroform on the PDMS soft mold. The films were dried at room temperature covered with
a glass dish. PLLA films were vacuum dried in a desiccator for 5 min to remove any residual solvent
before demolding. The PLLA films were trimmed to 1.0-cm
2
squares and sterilized with 70% ethanol
(v/v) and ultraviolet irradiation.
To verify the surface morphology and fidelity of the micro- and nano-topography replication process,
PLLA films were sputter coated with gold (JEOL JFC 1600 Fine Gold Coater, JEOL Ltd., Tokyo, Japan)
and examined by scanning electron microscopy (SEM, Quanta FEG 200 and JEOL JSM-5600LV
Scanning Microscope, JEOL Ltd., Tokyo, Japan).
5.2. Cell Culture of Mammary Epithelial Cells, MCF7 and Primary Invasive Ductal Carcinoma Cells
HMEC (Lonza, Basel, Switzerland) were maintained using mammary epithelial basal medium
(MEBM, Lonza) supplemented with bovine pituitary extract (BPE), human epithelial growth factor
(hEGF), hydrocortisone, GA-1000 (Gentamicin/Amphotericin) and insulin. The medium was changed
on alternate days. Cells were subcultured by washing with HEPES-buffered saline solution
(HEPES-BSS), dissociated with 0.25× trypsin-EDTA and neutralized with trypsin-neutralizing solution
J. Funct. Biomater. 2015, 6 254
(TNS) (Lonza). The human breast cancer cell line MCF7 was obtained from American Type Culture
Collection (ATCC) and maintained using Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma
Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS), 0.01 mg/mL bovine
insulin, 1% penicillin streptomycin and 1% non-essential amino acid solution. The medium was changed
every three days. Invasive ductal carcinoma (IDC) cells (Passage 3–5) from breast tumor tissues of
patient donors were obtained from the NUH-NUS (Singapore) Tissue Repository with NUS Institutional
Review Board approval (08-090). They were maintained using the mammary epithelial growth medium
(MEGM, Lonza) supplemented with 10% FBS, amphotericin B and gentamicin.
Cells were trypsinized upon reaching 80% confluence and seeded on PLLA films with a seeding
density of 10,000 cells/cm
2
(MCF7) or 16,000 cells/cm
2
(HMEC). The substrates were then placed in an
incubator for static culture. In the co-culture of MCF7 and HMEC, 16,000 cells/cm
2
of each cell type
(HMEC and MCF7) were seeded onto each of the samples and the control. The cell culture was
maintained using MEGM. To differentiate both cell types in the co-culture system, HMEC was stained
with CellTracker™ Red CMPTX (Invitrogen, Waltham, MA, USA) by incubating with 10 µM of the
dye in serum-free medium (MEGM) for 30 min and subsequently replaced with MEGM for another
30 min.
5.3. Cell Adhesion Counting
The number of attached cells on nanotopographic substrates in monocultures was determined after
4 and 24 h. Unattached cells were washed twice with PBS or HBSS. Cells were then detached with
trypsin from the PLLA films, and the cell number was determined by hemocytometer counting.
Adhesion assays were performed with 5 replicates.
In another method, cell adhesion counting was performed after 4 h of culture. In order to identify the
live cells, the attached cells were stained with the dye, CellTracker™ Red CMPTX (Invitrogen). After
3 h of culture, cells were labeled with CellTracker dye. At the 4-hour time point, the samples were gently
washed to remove the unattached cells, then fixed with 4% formaldehyde solution and stained with
DAPI. The samples were washed and mounted before imaging with a fluorescent microscope (Leica DC
300F, Leica, Wetzlar, Germany). The number of adhered cells was determined by counting the DAPI-
and CellTracker-stained cells. For each sample, 5 random images were taken at 20× magnification. Cell
adhesion counting was carried out for both HMEC and MCF7 mono-cultures.
5.4. Immunostaining of CD44, CD24 and Epithelial Surface Antigen Markers
After 4 or 24 h of culture, the cells were fixed with 4% formaldehyde and incubated in blocking buffer
(1% w/v bovine serum albumin (BSA) and 10% goat serum in PBS). The cells were stained with
antibodies specific for human cell surface markers: CD44-FITC (Abcam, Cambridge, UK), CD24-PE
(Abcam) and ESA (Santa Cruz, Starr County, TX, USA) (Table 1). After overnight incubation at 4 °C,
the cells were washed and incubated with the corresponding secondary antibodies: Alexa Fluor 488,
Alexa Fluor 546 and Alexa Fluor 647 (Invitrogen) for 1 h at room temperature. The cells were
counter-stained with DAPI. For each sample, 5 random images were taken at 20× magnification. The
number of cells that were of the CD44
+
CD24
−
ESA
+
phenotype was counted. The CD24
−/low
immunophenotype was taken as CD24
−
for cell counting.
J. Funct. Biomater. 2015, 6 255
Table 1. Dilution ratio of the antibodies used in immunostaining.
Antigen Primary antibody Secondary antibody
CD44 Anti-CD44, rat, conjugated with FITC 1:500 FITC, anti-rat 1:750
CD24 Anti-CD24, mouse, conjugated with PE 1:50 Alexa Fluor 546, anti-mouse 1:750
ESA Anti-ESA, rabbit, unconjugated 1:50 Alexa Fluor 647, anti-rabbit 1:750
5.5. Flow Cytometry Analysis
To analyze the expression of CD44, CD24 and ESA after 4 or 24 h of culture, flow cytometry was
performed on the MCF7 mono-culture and IDC cell experiment. At the 4- or 24-hour time point, the
cells attached on the films were trypsinized, centrifuged and counted. The harvested cells were incubated
with 2% goat serum in staining buffer (1% goat serum, 0.5% BSA in PBS) for 10 min in an ice bath and
stained with antibodies specific for human cell surface markers: CD44-FITC (Abcam), CD24-PE
(Abcam) and ESA (Santa Cruz) (Table 2). After incubation for 30 minutes at 4 °C, the suspensions were
washed three times in staining buffer. Cells were then incubated with the secondary antibody, Alexa
Fluor 647 for ESA, for 30 min at 4 °C. The cells were washed twice and fixed with 1% formaldehyde
for 1 h before analyzing with the flow cytometer (Dako Cytomation Cyan LX, Glostrup, Denmark).
The data were analyzed by FlowJo (version 10.0.7).
Table 2. Dilution ratio of the antibodies used in flow cytometry.
Antigen Primary antibody Secondary antibody
CD44 Rat anti-CD44 conjugated with FITC 1:50 None
CD24 Mouse anti-CD24 conjugated with PE 1:10 None
ESA Rabbit anti-ESA Un-conjugated 1:100 Goat anti-rabbit conjugated with Alexa Fluor 647 1:500
5.6. Statistics Analysis
Data are presented as the mean ± SD unless specified otherwise. The statistical significance of the
results obtained was assessed by one-way analysis of variance (ANOVA) and Tukey’s post hoc test or
the Kruskal–Wallis test followed by Dunn’s post hoc test for flow cytometry analysis data. Results with
P < 0.05 were considered statistically different.
Acknowledgments
This work is supported by the National University of Singapore Cross-Faculty Research Grant for
Interdisciplinary Research (XFac2008_P08) and the Singapore National Research Foundation under its
Research Centre of Excellence, administered by the Mechanobiology Institute, Singapore, at the
National University of Singapore. We would like to thank Soh-Zeom Yow, Lynn Lim, Terrence Kung
and Benjamin K.K. Teo for their excellent technical assistance in the experiments, as well as Dawn Neo,
Azita Gorji and Marie Cutiongco for proofreading the manuscript.
J. Funct. Biomater. 2015, 6 256
Author Contributions
All authors contributed to the design of the experiments. Christine S.Y. Giam and Ming Yi Leow
performed the in vitro experiments. Christine S.Y. Giam, Evelyn K.F. Yim and Kenneth K.B. Tan
performed data analysis and interpretation. Ching-Wan Chan provided scientific expertise.
Evelyn K.F. Yim and Kenneth K.B. Tan wrote the manuscript.
Conflicts of Interest
The authors declare no conflict of interest.
References
1. Visvader, J.E.; Lindeman, G.J. Cancer stem cells in solid tumours: Accumulating evidence and
unresolved questions. Nat. Rev. Cancer 2008, 8, 755–768.
2. Creighton, C.J.; Li, X.; Landis, M.; Dixon, J.M.; Neumeister, V.M.; Sjolund, A.; Rimm, D.L.;
Wong, H.; Rodriguez, A.; Herschkowitz, J.I.; et al. Residual breast cancers after conventional
therapy display mesenchymal as well as tumor-initiating features. Proc. Natl. Acad. Sci. USA 2009,
106, 13820–13825.
3. Lagadec, C.; Vlashi, E.; Della Donna, L.; Meng, Y.; Dekmezian, C.; Kim, K.; Pajonk, F. Survival
and self-renewing capacity of breast cancer initiating cells during fractionated radiation treatment.
Breast Cancer Res. 2010, 12, doi: 10.1186/bcr2479.
4. Al-Hajj, M.; Wicha, M.S.; Benito-Hernandez, A.; Morrison, S.J.; Clarke, M.F. Prospective
identification of tumorigenic breast cancer cells. Proc. Natl. Acad. Sci. USA 2003, 100, 3983–3988.
5. Abraham, B.K.; Fritz, P.; McClellan, M.; Hauptvogel, P.; Athelogou, M.; Brauch, H. Prevalence of
CD44
+
/CD24
−/low
cells in breast cancer may not be associated with clinical outcome but may favor
distant metastasis. Clin Cancer Res. 2005, 11, 1154–1159.
6. Ponti, D.; Costa, A.; Zaffaroni, N.; Pratesi, G.; Petrangolini, G.; Coradini, D.; Pilotti, S.;
Pierotti, M.A.; Daidone, M.G. Isolation and in vitro propagation of tumorigenic breast cancer cells
with stem/progenitor cell properties. Cancer Res. 2005, 65, 5506–5511.
7. Sheridan, C.; Kishimoto, H.; Fuchs, R.K.; Mehrotra, S.; Bhat-Nakshatri, P.; Turner, C.H.;
Goulet, R., Jr.; Badve, S.; Nakshatri, H. CD44
+
/CD24
−
breast cancer cells exhibit enhanced invasive
properties: An early step necessary for metastasis. Breast Cancer Res. 2006, 8, doi:10.1186/bcr1610.
8. Fillmore, C.M.; Kuperwasser, C. Human breast cancer cell lines contain stem-like cells that
self-renew, give rise to phenotypically diverse progeny and survive chemotherapy. Breast Cancer
Res. 2008, 10, doi: 10.1186/bcr1982.
9. Gires, O.; Klein, C.A.; Baeuerle, P.A. On the abundance of epcam on cancer stem cells. Nat. Rev.
Cancer 2009, 9, doi:10.1038/nrc2499-c1.
10. Osta, W.A.; Chen, Y.; Mikhitarian, K.; Mitas, M.; Salem, M.; Hannun, Y.A.; Cole, D.J.;
Gillanders, W.E. Epcam is overexpressed in breast cancer and is a potential target for breast cancer
gene therapy. Cancer Res. 2004, 64, 5818–5824.
11. Morrison, B.J.; Steel, J.C.; Morris, J.C. Sphere culture of murine lung cancer cell lines are enriched
with cancer initiating cells. PloS one 2012, 7, doi: 10.1371/journal.pone.0049752.
J. Funct. Biomater. 2015, 6 257
12. Wu, C.; Alman, B.A. Side population cells in human cancers. Cancer Lett. 2008, 268, 1–9.
13. Patrawala, L.; Calhoun, T.; Schneider-Broussard, R.; Zhou, J.; Claypool, K.; Tang, D.G. Side
population is enriched in tumorigenic, stem-like cancer cells, whereas ABCG2+ and ABCG2−
cancer cells are similarly tumorigenic. Cancer Res. 2005, 65, 6207–6219.
14. Voorhees, P.M.; Dees, E.C.; O’Neil, B.; Orlowski, R.Z. The proteasome as a target for cancer
therapy. Clin. Cancer Res. 2003, 9, 6316–6325.
15. Vlashi, E.; Kim, K.; Lagadec, C.; Donna, L.D.; McDonald, J.T.; Eghbali, M.; Sayre, J.W.;
Stefani, E.; McBride, W.; Pajonk, F. In vivo imaging, tracking, and targeting of cancer stem cells.
J. Natl. Cancer Inst. 2009, 101, 350–359.
16. Martinez, E.; Engel, E.; Planell, J.A.; Samitier, J. Effects of artificial micro- and nano-structured
surfaces on cell behaviour. Ann Anat. 2009, 191, 126–135.
17. Yim, E.K.; Sheetz, M.P. Force-dependent cell signaling in stem cell differentiation. Stem Cell Res.
Ther. 2012, 3, doi: 10.1186/scrt132.
18. Teo, B.K.; Ankam, S.; Chan, L.Y.; Yim, E.K. Nanotopography/mechanical induction of stem-cell
differentiation. Methods Cell Biol. 2010, 98, 241–294.
19. Kwon, K.W.; Lee, S.H.; Kim, B.; Lee, S.N.; Park, M.C.; Kim, P.; Hwang, S.Y.; Suh, K.Y.
Label-free, microfluidic separation and enrichment of human breast cancer cells by adhesion
difference. Lab Chip 2007, 7, 1461–1468.
20. Radisic, M.; Iyer, R.K.; Murthy, S.K. Micro- and nano-technology in cell separation. Int. J.
Nanomed. 2006, 1, 3–14.
21. Yim, E.K.; Leong, K.W. Significance of synthetic nanostructures in dictating cellular response.
Nanomedicine 2005, 1, 10–21.
22. Sagar, J.; Chaib, B.; Sales, K.; Winslet, M.; Seifalian, A. Role of stem cells in cancer therapy and
cancer stem cells: A review. Cancer Stem Cell Int. 2007, 7, doi:10.1186/1475-2867-7-9.
23. Curtis, A.; Wilkinson, C.Topographical control of cells. Biomaterials 1997, 18, 573–583.
24. Park, C.; Bergsagel, D.E.; McCulloch, E.A. Mouse myeloma tumor stem cells: A primary cell
culture assay. J. Natl. Cancer Inst. 1971, 46, 411–422.
25. Ponti, D.; Zaffaroni, N.; Capelli, C.; Daidone, M.G. Breast cancer stem cells: An overview.
Eur. J. Cancer 2006, 42, 1219–1224.
26. Dick, J.E. Breast cancer stem cells revealed. Proc. Natl. Acad. Sci. USA 2003, 100, 3547–3549.
27. Shipitsin, M.; Campbell, L.L.; Argani, P.; Weremowicz, S.; Bloushtain-Qimron, N.; Yao, J.;
Nikolskaya, T.; Serebryiskaya, T.; Beroukhim, R.; Hu, M. Molecular definition of breast tumor
heterogeneity. Cancer Cell 2007, 11, 259–273.
28. Al-Hajj, M.; Becker, M.W.; Wicha, M.; Weissman, I.; Clarke, M.F. Therapeutic Applications
Cancer Stem Cells; Elsevier: Amsterdam, The Netherlands, 2004; pp. 43–47.
29. Lostumbo, A.; Mehta, D.; Setty, S.; Nunez, R. Flow cytometry: A new approach for the molecular
profiling of breast cancer. Exp. Mol. Pathol. 2006, 80, 46–53.
30. Plouffe, B.D.; Njoka, D.N.; Harris, J.; Liao, J.; Horick, N.K.; Radisic, M.; Murthy, S.K.
Peptide-mediated selective adhesion of smooth muscle and endothelial cells in microfluidic shear
flow. Langmuir 2007, 23, 5050–5055.
J. Funct. Biomater. 2015, 6 258
31. Dalby, M.J.; Riehle, M.O.; Sutherland, D.S.; Agheli, H.; Curtis, A.S. Use of nanotopography to
study mechanotransduction in fibroblasts—Methods and perspectives. Eur. J. Cell Biol. 2004, 83,
159–169.
32. Dalby, M.J. Topographically induced direct cell mechanotransduction. Med. Eng. Phys. 2005, 27,
730–742.
33. Ingber, D.E. Tensegrity I. Cell structure and hierarchical systems biology. J. Cell Sci. 2003, 116,
1157–1173.
34. Ingber, D.E. Tensegrity II. How structural networks influence cellular information processing
networks. J. Cell Sci. 2003, 116, 1397–1408.
35. Teo, B.K.; Wong, S.T.; Lim, C.K.; Kung, T.Y.; Yap, C.H.; Ramagopal, Y.; Romer, L.H.; Yim, E.K.
Nanotopography modulates mechanotransduction of stem cells and induces differentiation through
focal adhesion kinase. ACS Nano 2013, 7, 4785–4798.
36. Wong, S.T.; Teo, S.K.; Park, S.; Chiam, K.H.; Yim, E.K. Anisotropic rigidity sensing on grating
topography directs human mesenchymal stem cell elongation. Biomech. Model. Mechanobiol. 2014,
13, 27–39.
37. Ankam, S.; Lim, C.K.; Yim, E.K.F. Actomyosin contractility plays a role in MAP2 expression
during nanotopography-directed neuronal differentiation of human embryonic stem cells.
Biomaterials 2015, 47, 20–28.
38. Teo, B.K.K.; Goh, K.J.; Ng, Z.J.; Koo, S.; Yim, E.K.F. Functional reconstruction of corneal
endothelium using nanotopography for tissue-engineering applications. Acta Biomater. 2012, 8,
2941–2952.
© 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/4.0/).
