Biomaterials 69 (2015) 121e132

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Cell type-specic adaptation of cellular and nuclear volume in

micro-engineered 3D environments
Alexandra M. Greiner a, Franziska Klein a, Tetyana Gudzenko b, Benjamin Richter a,
Thomas Striebel a, Bayu G. Wundari a, Tatjana J. Autenrieth c, Martin Wegener b, d,
Clemens M. Franz b, *, Martin Bastmeyer a, b, c, **
a

Karlsruhe Institute of Technology (KIT), Zoological Institute, Department of Cell- and Neurobiology, Haid-und-Neu-Strae 9, D-76131 Karlsruhe, Germany
Karlsruhe Institute of Technology (KIT), DFG-Center for Functional Nanostructures (CFN), Wolfgang-Gaede-Strae 1a, D-76131 Karlsruhe, Germany
Karlsruhe Institute of Technology (KIT), Institute of Functional Interfaces (IFG), 76344 Eggenstein-Leopoldshafen, Germany
d
Karlsruhe Institute of Technology (KIT), Institute for Applied Physics, Wolfgang-Gaede-Strae 1, D-76131 Karlsruhe, Germany
b
c

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 5 August 2015
Accepted 7 August 2015
Available online 8 August 2015

Bio-functionalized three-dimensional (3D) structures fabricated by direct laser writing (DLW) are
structurally and mechanically well-dened and ideal for systematically investigating the inuence of
three-dimensionality and substrate stiffness on cell behavior. Here, we show that different broblast-like
and epithelial cell lines maintain normal proliferation rates and form functional cell-matrix contacts in
DLW-fabricated 3D scaffolds of different mechanics and geometry. Furthermore, the molecular composition of cell-matrix contacts forming in these 3D micro-environments and under conventional 2D
culture conditions is identical, based on the analysis of several marker proteins (paxillin, phosphopaxillin, phospho-focal adhesion kinase, vinculin, b1-integrin). However, broblast-like and epithelial
cells differ markedly in the way they adapt their total cell and nuclear volumes in 3D environments.
While broblast-like cell lines display signicantly increased cell and nuclear volumes in 3D substrates
compared to 2D substrates, epithelial cells retain similar cell and nuclear volumes in 2D and 3D environments. Despite differential cell volume regulation between broblasts and epithelial cells in 3D environments, the nucleus-to-cell (N/C) volume ratios remain constant for all cell types and culture
conditions. Thus, changes in cell and nuclear volume during the transition from 2D to 3D environments
are strongly cell type-dependent, but independent of scaffold stiffness, while cells maintain the N/C ratio
regardless of culture conditions.
2015 Elsevier Ltd. All rights reserved.

Keywords:
Direct laser writing
3D cell culture substrates
Cell-matrix adhesions
Cell volume
Nuclear volume

1. Introduction
Much of the current knowledge of adherent cell behavior has
been obtained by in vitro cell culture experiments using conventional stiff and planar plastic or glass substrates. These twodimensional (2D) cell culture substrates have been improved by
micro-contact printing techniques, resulting in uniform [1e3] or
graded patterns of adhesion molecules [4], or by micro-well

* Corresponding author. Karlsruhe Institute of Technology (KIT), DFG-Center for

Functional Nanostructures (CFN), Wolfgang-Gaede-Strae 1a, D-76131 Karlsruhe,
Germany.
** Corresponding author. Karlsruhe Institute of Technology (KIT), Zoological
Institute, Department of Cell- and Neurobiology, Haid-und-Neu-Strae 9, D-76131
Karlsruhe, Germany.
E-mail addresses: clemens.franz@kit.edu (C.M. Franz), bastmeyer@kit.edu
(M. Bastmeyer).
http://dx.doi.org/10.1016/j.biomaterials.2015.08.016
0142-9612/ 2015 Elsevier Ltd. All rights reserved.

structuring methods providing topographic cues [5e8]. Nevertheless, 2D cell culture conditions do not reect the three-dimensional
(3D) environment of cells in vivo. In contrast to stiff 2D cell culture
substrates, the natural extracellular matrix (ECM) is a compliant,
complex and information-rich 3D environment [9,10]. A growing
understanding of the shortcomings of 2D cell culture systems in
biomedical sciences has stimulated the development of novel, more
physiological 3D cell culture systems. These systems allow for
investigating factors that are operative only in 3D microenvironments, for instance guiding cellular phenomena in development, tissue homeostasis, and disease [10,11]. Depending on the
makeup of these scaffolds, cells show dramatic changes in cell
behavior compared to 2D culture [12]. For instance, cells cultured in
cell-derived 3D ECM scaffolds typically exhibit differences in
migration [13e15], proliferation behavior [16], or bronectin
brillogenesis [17]. Additionally, relatively little is known about the

122

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

structure and function of cell-matrix adhesions formed in 3D environments compared to the 2D situation. Cell-matrix adhesions
are important cellular sensors of matrix rigidity as they mechanically link the ECM with the acto-myosin system [9,10,18]. Reports
about cell-matrix adhesions in 3D vary widely across the literature
and are often in apparent conict [10,11,16,19e23]. It has been reported that cells differ in adhesion, adhesive signaling and overall
migration mode not only between 2D and 3D micro-environments,
but also between different 3D substrate types [11]. It has also been
shown that cells in soft cell-derived 3D matrices or 3D collagen
matrices develop so-called 3D cell-matrix adhesions, which differ
in molecular composition and localization, and morphology from
contacts formed on planar 2D substrates [16,19e26]. In addition,
the local ECM architecture, including the orientation and diameter
of ECM bers and consequently the functional area available for
cell-matrix adhesion formation, can inuence the appearance and
morphology of cell-matrix contacts [11,13,16,27].
Although naturally-derived 3D matrices retain important features of in vivo environments, their complex molecular composition, variable architecture (pore size, connement, degree of crosslinking) and variable stiffness complicates the systematic analysis
of their inuence on cell behavior [9]. To better understand cell
behavior in 3D environments, reductionist approaches are needed
to correlate cellular responses to particular chemical and physical
properties of the 3D substrate. To this end, a number of articial 3D
systems, including micro-carriers [28,29], synthetic hydrogels
[30,31], and micro-well arrays [5e7] have been established. These
systems offer reproducible biochemical conditions and some of
them can even mimic ECM porosity, but they frequently lack the
architectural, mechanical, and biochemical versatility necessary for
comprehensive cell-biological studies.
To overcome some of the limitations of articial planar and rigid
cell culture substrate, we [32e36] and others [12,28,37e40] have
used the direct laser writing (DLW) photo-polymerization technique
to fabricate tailored 3D scaffolds for single cell cultivation. By using
polymers with different mechanical properties or by adjusting
scaffold feature sizes, the stiffness of DLW-produced cellular microenvironments can be accurately adjusted [35]. Furthermore, DLW
scaffolds can be selectively bio-functionalized by sequential photoactivation [33,34]. Thus, DLW-produced 3D scaffolds can ll a gap
between the vast structural complexities of natural 3D cell environments on the one hand and the oversimplied 2D situation of
planar cell culture substrates for in vitro studies on the other hand.
In this study we demonstrate that bronectin-coated 3D scaffolds fabricated by DLW display different degrees of stiffness and
geometries and support cell proliferation in a similar manner as
standard 2D cell culture techniques. Furthermore, using different
cell-matrix adhesion marker proteins we show that the composition of cell-matrix adhesions in the 3D scaffolds is similar to those
formed on planar 2D cell culture surfaces, even though some
morphological differences were detected. However, we nd striking cell-type dependent differences in cell and nuclear volume
regulation between 2D and 3D culture conditions, but nuclear-tocell (N/C) volume ratios are maintained in 2D and 3D environments for all cell types. Tailored 3D environments produced by
DLW can therefore reveal cell type-specic changes in cell and
nuclear volume regulation, which may reect different tissuespecic tasks of these cell types in vivo.
2. Materials and methods
2.1. Direct laser writing
Direct laser writing (DLW) is a two photon polymerization (2 PP)
technique [35,36,41] in which a femto-second laser beam is focused

into a photo-sensitive liquid material consisting of a mixture of

monomeric matrix molecules and a photo-initiator. Simultaneous
absorption of two photons leads to the excitation of photo-initiator
molecules, which then trigger a highly localized chemical polymerization event that is conned to the focal volume of the laser
due to the non-linearity of the 2 PP process (Supplemental Fig S1)
[12,28,40,42]. The spot size for the 2 PP is diffraction limited and in
our setup has a size of approx. axy 314 nm and az 767 nm (for
l 780 nm, NA 1.4, n 1.52) [40]. For scaffold writing, a single
droplet (~100 ml) of the photo-sensitive viscous liquid material is
rst drop-cast onto a planar supporting carrier substrate. Due to its
high surface tension, the viscous photo-sensitive liquid retains a
roughly spherical shape on the carrier substrate. In our study the
carrier substrates were plasma-cleaned high precision microscopy
glass cover slips (22 mm  22 mm, 170 5 mm, No. 1.5H LH24.1
from Roth) functionalized with 3-(trimethoxysilyl) propyl methacrylate (Sigma Aldrich) dissolved at a concentration of 1 mM in
toluene (Sigma Aldrich). However, other transparent or nontransparent carrier substrates of different materials, e.g., polymers or silicon, are also suitable for the DLW fabrication process.
The writing process starts with focusing the laser on the interface
between the glass substrate and the photo-resist to ensure a stable
and covalent attachment of the emerging 3D polymer scaffold to
the carrier substrate [12]. After the writing process the substrates
were developed in a 1:1 mixture of methyl isobutyl ketone and isopropanol, rinsed in iso-propanol and dried in a nitrogen ow.
2.1.1. Stiff 3D PETTA structures
We have previously shown that DLW-induced polymerization of
a mixture of petaerythriol tetraacrylate (PETTA, SigmaeAldrich)
monomer and an appropriate photo-initiator produces 3D structures with increased physical stability and high protein adsorbing
properties [36]. The monomeric PETTA solution also contains about
400 ppm monomethyl ether hydroquinone to inhibit prepolymerization, but no other solvents. For DLW, one weight
percent of the monomeric PETTA was mixed with three weight
percent of Irgacure 379 photo-initiator (BASF). Typical writing parameters of the DLW system for this resist were an average laser
power of 20 mW and a writing speed of 200 mm/s. The produced
wheel structures have a total diameter of 102 mm. The spokes have a
height of 15 mm; the two sets of arcs which form the outer wall of
the wheel structure are each approx. 5 mm high and 1 mm thick. The
spokes are arranged at 36 angles to each other. The dimensions of
the outer and inner sector areas of the wheel are 582.92 mm2 and
185.64 mm2, respectively.
2.1.2. Soft 3D Ormocomp structures
Ormocomp (Micro Resist Technology, Berlin, Germany) is a
member of the Ormocer family of inorganic (SieOeSi)eorganic
hybrid polymers. It has been previously used to fabricate DLWproduced elastic scaffolds for single cell force measurements [35]
and for controlled 3D cell culture applications [36]. Ormocomp
possess good protein binding properties in its polymerized state.
Ormocomp is a proprietary viscous liquid containing an undisclosed photo-initiator (according to the manufacturer's information) and requires no further supplements for photopolymerization. During DLW the Ormocomp photo-resist was
polymerized at a laser speed of 200 mm/s and a minimal laser power of 12.5 mW, which produced beams with a cross section of
approximately 1.5 mm. The beams are arranged at 90 angles to each
other. The height of the pillars supporting the net structures is
20 mm and the spacing between the pillars varies incrementally
between 20 mm and 30 mm. The area of the squares formed by the
beams of a net is 400 mm2 and 900 mm2.

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

2.1.3. 2D substrates
As 2D substrates we used conventional glass cover slips
homogenously coated with human plasma bronectin (SigmaeAldrich, see below for details about the substrate coating). In
control experiments we also produced 2D PETTA and 2D Ormocomp substrates on glass cover slips using the same DLW setup as
for the corresponding 3D scaffolds (see above). However, the
writing parameters were adjusted to produce homogenous PETTA
or Ormocomp polymer lms of approximately 1 mm thickness
which were then coated with bronectin.
2.2. Cell culture
Buffalo rat liver cells (BRL, broblasts), mouse embryonic broblasts (MEF), and normal rat kidney cells (NRK, epithelial) were
purchased from the American Type Tissue Culture Collection (ATCC,
Manassas, VA, USA). Spindle-shaped B16F1 (B16) (mouse melanoma) cells were a generous gift of Beat Imhof (CMU, Genova) and
epithelial A549 cells (human lung carcinoma cells) were a generous
gift of Harald Kruge (ITG, KIT, Karlsruhe). BRL, B16, MEF, and NRK
cells were cultured in Dulbecco's Modied Eagle Medium (DMEM,
Gibco) and A549 cells F-12K medium (Gibco). Both cell culture
media were supplemented with 10% fetal calf serum (FCS,
HyClone). Cells were kept in an incubator at 37  C with 5% CO2 in a
humidied environment and passaged approximately two to three
times a week. Prior of experiments, cells were grown to 70e80%
conuence and detached from the cell culture container by trypsin/
EDTA (0.25% trypsin/1 mM ethylenediaminotetraacetate, Invitrogen). Cell detachment was stopped by adding DMEM medium
supplemented with 10% FCS. The cell suspension was centrifuged
and the resulting supernatant discarded. Cells were resuspended in
fresh medium and the cell numbers per ml were determined in a
Neubauer counting chamber. All cells were used for experiments up
to a passage number below 40.
2.3. Substrate coating and cell cultivation in 2D and 3D
The cell culture substrates (i.e. unmodied glass cover slips,
PETTA- or Ormocomp-covered glass cover slips, or glass cover slips
carrying 3D structures) were placed into a single well of a six-well
plate without further xation to the bottom of the well. All cover
slips were sterilized with 70% ethanol (Roth) and washed three
times with phosphate buffered saline (PBS, Gibco). The 3D structures and the 2D control substrates were coated with 10 mg/ml
bronectin at RT for 60 min and then rinsed with PBS twice to
remove unbound bronectin. A total of 50.000 cells in 3 ml of the
respective cell culture media were added per well and cultured in
serum-containing medium on the different 3D structures or planar
2D control substrates for three hours. For Matrigel (BD Bioscience)
experiments cells suspended in PBS were carefully mixed with
Matrigel (nal Matrigel concentration 80%) and applied to a planar
glass surfaces. The Matrigel was left to stiffen for 15 min at 37  C
before cells were cultured for three hours in serum-containing
medium under standard cell culture conditions.
2.4. Immuno-uorescence
Cells on 2D surfaces or in 3D structures were xed for 10 min
with 4% paraformaldehyde (SigmaeAldrich) in PBS. After permeabilization in 0.1% Triton/PBS (SigmaeAldrich), cells were
incubated with primary antibodies in 1% bovine serum albumin
(SigmaeAldrich) in PBS for 1 h at RT: (a) mouse monoclonal anti
phospho-tyrosine (Tyr99) (1:100) (Santa Cruz, Heidelberg); (b)
mouse monoclonal anti focal adhesion kinase (FAK) (1:100) (BD
Bioscience); (c) rabbit polyclonal anti phospho-FAK (Tyr397)

123

(1:300) (Biosource, Solingen); (d) mouse monoclonal anti paxillin

(1:500) (BD Bioscience); (e) rabbit polyclonal anti phospho-paxillin
(Tyr118) (1:400) (Cell Signaling, Danvers/MA, USA), (f) rabbit
polyclonal anti bronectin (1:400) (SigmaeAldrich). After thorough rinsing with PBS, the samples were incubated with secondary
antibodies and dyes in 1% BSA/PBS for 1 h at RT: (i) goat anti mouse
Cy3 (1:500) (Dianova, Hamburg); (ii) goat anti rabbit Cy3 (1:500)
(Dianova); (iii) goat anti rabbit AlexaFluor488 (1:200, Invitrogen);
(iv) DAPI (1:1000) (SigmaeAldrich); (v) phalloidin AlexaFluor488
(1:200, Invitrogen). Samples were embedded in mounting media
(Mowiol, Hoechst) containing 1% N-propyl-gallate as antifade
(SigmaeAldrich). Matrigel-embedded cells were xed for 15 min
with 4% paraformaldehyde/PBS at 37  C. After permeabilization
with 0.1% Triton/PBS, cells were incubated with a mouse monoclonal paxillin antibody (1:500, BD Bioscience) in 1% BSA/PBS
overnight at RT. After thorough rinsing with PBS, samples were
incubated with the corresponding secondary antibodies and dyes
in 1% BSA/PBS overnight at RT (goat anti mouse Cy3 (1:500) (Dianova); DAPI (1:500); phalloidin AlexaFluor488 (1:200)).
2.5. Microscopy and image analysis
Cells in 3D structures embedded in mounting media were
analyzed using a laser scanning microscope (LSM 510 Meta, Zeiss)
equipped with a Plan Apochromat 20/0.8 objective (Zeiss) and LCI
Plan Neouar 63/1.3 DIC ImKorr objective (Zeiss). Maximum intensity projections were produced using the LSM Zeiss software. 3D
reconstructions of the confocal image stacks were prepared with
the Volocity Software version 4.3.2 (Perkin Elmer). For atomic force
microscopy (AFM) imaging, cells on planar 2D control substrates
were transferred into 35 mm cell culture dishes and imaged in PBS
using a NanoWizard II AFM (JPK Instruments, Germany) mounted
on top of an inverted optical microscope (Carl Zeiss AxioObserver
A1) and a designated Petri dish holder. AFM imaging was performed in contact mode at a typical scan frequency of 0.4 Hz with
V-shaped cantilevers with a spring constant of 0.03 N/m (MLCT,
Bruker). For scanning electron microscopy (SEM), scaffold samples
required no dehydration and were sputtered with gold (6e10 nm)
before they were imaged under vacuum at 5 kV using a Supra 55
scanning electron microscope (Zeiss).
2.6. Cell volume measurements in 2D and 3D micro-environments
Cell volumes in 2D were calculated from AFM height images
using two independent software systems: (1) Gwyddion (http://
gwyddion.net/) and (2) Matlab (Mathworks). When using the
Gwyddion software, the base level of the AFM height images was
determined by placing a plane through three ctive points positioned at the cells-substrate border. A height threshold was then
adjusted to isolate the cell from the substrate and to obtain the cell
volume data. Volume calculation using Matlab was performed using AFM height image in ASCII text le format. The cell volume was
then calculated using the following formula:

dV

z  dA

z  dx  dy

(1)

where z is the height of a data point, dA the basal plane which is the
product of the distance of two pixels in horizontal and vertical
direction. The total volume corresponds to the sum of the single
volume elements. The zero level was set manually and the average,
minimum and maximum values were determined. The cell volume
was then calculated from the difference of the measured value to
the average, minimum and maximum value.
To obtain cell volumes in 2D and 3D by optical microscopy,

124

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

samples were stained for F-actin and imaged by confocal laser

scanning microscopy (CLSM). Complete image z-stacks of single
cells were generated with the distance between two neighboring zslices set to 410 nm. For volume analysis, the cell cross section area
(mm2) in each z-slice of the confocal image stack was rst determined using morphometric segment analysis in Metamorph
(Visitron). The cross section areas from all z-slice were then added,
and the sum of the single area elements multiplied with the zspacing value (here 410 nm) to obtain the volume (mm3) of a single
cell. In the case of the 3D structures, only cells that were fully
suspended in the third dimension were evaluated and only cells
that adhered to the outer segment of the 3D wheel structure were
analyzed.
2.7. Nuclear volume measurements in 2D and 3D microenvironments
The nuclear volume of DAPI-stained cells was determined from
CLSM image z-stacks using a custom-made Matlab routine. The
CLSM image intensity was normalized to a 0e1 range according to
the following equation:

Inormalized



Iinput emin Iinput




max Iinput  min Iinput

(2)

where min(Iinput) and max(Iinput) denotes the minimum and

maximum intensity of the input image, respectively. Image
denoising based on wavelet (Bayes soft thresholding) and Wiener
lter was performed on the normalized CLSM image. This procedure was followed by image deconvolution based on the scaled
gradient projection algorithm [43]. To determine the boundary of
the cell nucleus, level-set based image segmentation was performed on the deconvoluted image [44]. The resulting binary image
was furthered processed by image dilation and erosion with onepixel-sized diamond structure element. The nucleus volume was
then computed from the binary image using Minkowski measures
[45]. Nuclear-to-cell volume ratios (N/C ratio) were obtained by
dividing the nuclear volume by the cell volume.
2.8. Statistics
OriginPro 8G software (OriginLab Cooperation, Northampton,
USA) was used for statistical analysis (ANOVA test). Differences
were considered as statistically signicant when the calculated p
value was less than 0.05 (*). A p value above 0.05 was considered
non-signicant (z, ).
3. Results
3.1. Tailored bio-functionalized 3D cell culture substrates support
cell proliferation
The majority of cell culture experiments are performed on rigid
and planar (2D) substrates, even though in vivo most cells reside in
pliable 3D environments. New technologies to fabricate dened 3D
culture substrates are becoming increasingly available, but the
impact of 3D cell culture conditions on single cell behavior is still
incompletely understood. To analyze the inuence of 3D culture
conditions on cell behavior systematically, we used DLW to produce
two types of 3D polymeric scaffolds with different mechanical
properties. The rst scaffold type is made of pentaerythritol tetraacrylate (PETTA) and contains attened beam structures arranged
into a wheel-like architecture (Fig. 1A). This conguration results in
a mechanically stable scaffold that cannot be visibly deformed by

cellular contraction forces (stiff scaffold). In the second scaffold

type made of Ormocomp, pillars are connected by thin, exible
beams (Fig. 1B), yielding a scaffold that can be easily deformed by
contractile forces exerted by single cells (soft scaffold). Stiff and
soft thus refer to the ability (or inability) of a single cell to deform
parts of these scaffolds. Both PETTA and Ormocomp have comparable E-moduli values in the high three-digit MPa or even GPa
range. Scaffold deformability thus depends primarily on its geometry, rather than the E-modulus of the polymeric material.
We rst tested the biocompatibility of the produced 3D DLWscaffolds and analyzed whether they supported fundamental cell
functions, such as cell adhesion, growth, and proliferation. Both
scaffold types were homogeneously coated with bronectin to
support stable cell attachment and to provide identical surface
chemistry (Supplemental Fig. S2). Buffalo rat liver broblasts (BRL)
grown in these 3D micro-environments easily adapted their cell
shape to the complex scaffold geometries (Fig. 2), indicating efcient attachment to both scaffold types. Proliferation rates of BRL
cells were similar in both 3D scaffold types and when compared to
2D control cell culture substrates (bronectin-coated glass cover
slips, Supplemental Fig. S3). This indicates that both 3D scaffold
types support normal cell proliferation rates.
3.2. Comparing cell-matrix adhesions between 2D and 3D microenvironments
Previous studies have shown that cells cultured in cell-derived
3D matrices develop cell-matrix adhesions which differ in
morphology and molecular composition from those formed on
planar 2D substrates [16,23,24]. To test whether the dimensionality
and mechanics of our scaffolds affect the formation of 3D cellmatrix adhesions, we analyzed the appearance of prominent
adhesion markers by immuno-uorescence staining. For this, we
cultured BRL cells on bronectin-coated 2D control surfaces and in
stiff or soft 3D scaffolds and rst visualized paxillin-containing cellmatrix adhesions (Fig. 3AeC). Paxillin localized into classical focal
adhesions on 2D substrates or into elongated patches along the
beams of the 3D scaffolds (Fig. 3B, C). In addition, the prominent
cell-matrix adhesion markers phospho-tyrosine (Tyr99), phosphofocal adhesion kinase (pFAK) (Tyr397), phospho-paxillin (Tyr118)
were present in all cells cultured in the 3D scaffolds, similar to cells
growing on planar glass substrates (Supplemental Fig. S4). However, we noticed a slight reduction of FAK and phospho-FAK
staining intensity in both 3D scaffold types compared to 2D substrates. Furthermore, cell-matrix adhesion sites also stained for
vinculin and b1-integrin in BRLs cultured in 3D scaffolds (data not
shown). In addition, BRLs cultured in DLW-fabricated 3D scaffold
formed pronounced actin stress bers primarily located along the
cell periphery (Fig. 3B, C). These results established that
bronectin-coated 3D scaffolds (independently of their mechanics
and architecture) efciently promote cell-matrix adhesion site
formation similar to conventional 2D culture conditions.
3.3. Cell and nuclear volume varies between 2D and 3D microenvironments in a cell-type dependent manner
To investigate potential cell type-specic effects of 3D culture,
we analyzed overall cell morphology in several mesenchymal and
epithelial cell lines. As further examples of mesenchymal cells in
addition to BRL cells, we studied mouse embryonic broblasts
(MEF) and B16 cells from a mouse melanoma tumor. As examples
for epithelial cells we investigated A549 cells from human lung
carcinoma and normal rat kidney cells (NRK). The different cell
types were cultured on bronectin-coated 2D glass surfaces and in
stiff 3D scaffolds for three hours. We then analyzed the morphology

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

125

Fig. 1. Tailored 3D scaffolds fabricated by direct laser writing (DLW). DLW is a single step technique where a femto-second laser beam is focused into a photo-sensitive liquid
material. In the focal volume of the focused laser beam photo-initiator molecules are excited and cause a highly localized chemical polymerization event. (A) Scanning electron
microscopy (SEM) image of a 3D stiff scaffold made of pentaerythritol tetraacrylate (PETTA). (B) SEM image of a 3D soft scaffold made of Ormocomp and composed of pillars
interconnected with beams.

of all cell types after immuno-staining (Fig. 4A for a merged projection, Fig. 4B for a 3D reconstruction). All cell types attached to
the 3D scaffolds, but they displayed marked morphological differences. Epithelial cells were often in close contact with the walls of
the 3D structure and elongated and roughly planar in morphology.
In contrast, broblasts typically explored the full 3D environment,
spanning their cell body across different 3D segments, and displayed an angular and bulky morphology. The extensive spreading
behavior suggested that broblasts display increased cell volumes
in 3D environments.
To quantify cell volume in both 2D and 3D, we developed a dual
approach using AFM and CLSM imaging. Cell volume measurements of thin, well-spread cells on 2D substrates were performed
by AFM height imaging in liquid (Fig. 4C). AFM imaging generates
high resolution topographic images of surface-attached cells and
provides the highest volume accuracy for 2D samples, but is typically limited to samples displaying comparatively low height
changes. Since cells spreading on 2D surfaces are usually at
(~1e3 mm), they could be easily imaged by AFM (Supplemental
Fig. S5). In contrast, CLSM imaging is diffraction-limited and can
only offer a realistic resolution of ~500 nm in z-direction. In addition, given the low height of well-spread cells, only a few confocal
slices will intersect the cell, potentially decreasing the accuracy of
the volume data further. CLSM is therefore less suited for determining cellular volumes in 2D. To nevertheless determine the
correlation between AFM and CLSM 2D volume data given the
specic pros and cons of each technique, we cultured BRL cells on
bronectin-coated glass cover slips and rst determined the volumes of individual cells by AFM imaging (Supplemental Fig. S5).
Subsequently, we stained the samples for F-actin and imaged the
respective cells by CLSM. Thus, we generated both AFM and CLSM
volume data for each individual cell. The cell volumes determined
by AFM and CLSM correlated well for individual cells and were not
signicantly different when averaged over all analyzed cells,
demonstrating the general feasibility of both techniques for
determining cellular volumes. However, cell volumes determined
by CLSM were systematically higher by 10e15% compared to cell
volumes determined by AFM. We interpreted the slightly higher
CLSM volumes to reect optical diffraction effects, leading to a
slight broadening of the cell contour and overestimation of the true
cellular volume from confocal stacks. We considered the AFM data
to provide a more accurate account of cell volumes and subsequently used AFM volume data in the 2D system in our analysis.
However, AFM is a surface scanning technique incompatible with
complex 3D samples. In these cases, analysis of CLSM z-stacks
provided a good estimate of cellular volumes, especially since the

often greatly increased cellular extension in z-direction could be

sampled more accurately with a larger number of confocal slices.
The complementary AFM and CLSM analysis conrmed that all
broblast-like cell lines display signicantly larger cell volumes
when cultured in stiff 3D scaffolds compared to planar 2D surfaces
(Fig. 4D, BRL: single cell volume in 2D 878 mm3, ~55% increase in
volume in 3D; B16: 1586 mm3 in 2D, ~27% increase in volume in 3D;
MEF: 591 mm3 in 2D, ~63% increase in volume in 3D; ANOVA,
*p < 0.05). In contrast to broblast-like cell lines, the epithelial cell
lines A549 and NRK displayed similar cell volumes when growing
in bronectin-coated 3D scaffolds or on planar 2D surfaces (Fig. 4D,
A549 cells 2D: 1675 mm3, 3D: 1519 mm3; NRK cells 2D: 681 mm3 and
3D: 831 mm3; ANOVA, z, p > 0.05). These results thus revealed a
striking difference in cell volume regulation between broblastlike and epithelial cells in 3D micro-environments. Additional 2D
control experiments demonstrated that cell volume regulation is
independent of the underlying polymer chemistry of the cell culture substrates. The volume of BRL broblasts cultured at 2D conditions was similar on bronectin-coated glass cover slips or thin
bronectin-coated PETTA or Ormocomp lms (Supplemental
Fig. S5). Hence, BRL cell volumes were generally smaller on all 2D
substrates compared to 3D cell culture conditions (compare Fig. 4D
with Supplemental Fig. S5).
We furthermore assessed whether BRL cell volume differs in 3D
micro-environments of different mechanics and geometries.
However, BRL cells cultured in any 3D micro-environment fabricated by DLW displayed similar cell volumes (Figs. 5 and 3D stiff:
1942 mm3, 3D soft: 2066 mm3; ANOVA, z, p > 0.05). Scaffold stiffness
and geometry therefore appear to be irrelevant for cell volume
regulation. However, natural ECMs have a heterogeneous composition, while our 3D scaffolds were only functionalized with a single
ECM component (bronectin). We therefore also used Matrigel, a
soft, gel-like, non-brillar cell-derived heterogeneous protein
mixture primarily composed of laminin and collagen IV and containing different growth factors to introduce an alternative, more
complex 3D environment mimicking the in vivo ECM-environment
of cells. After culturing BRLs for three hours in pure Matrigel, the
cell volume was similar to BRLs growing in either stiff and soft
DLW-fabricated 3D scaffold types (Fig. 5, Matrigel: 1977 mm3;
ANOVA, z, p > 0.05; Supplemental Fig. S6). Thus, BRL cell volume
was signicantly larger in all three 3D culture conditions (wheels,
nets, Matrigel) - regardless of scaffold geometry and stiffness e
then on 2D control substrates (Figs. 5 and 2D: 878 mm3; ANOVA,
*p < 0.05).
We also assessed the nuclear volume of broblast and epithelial
cell lines on 2D substrates and in stiff 3D scaffolds based on CLSM-

126

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

Fig. 2. Growth of cells in stiff and soft 3D scaffolds. Buffalo rat liver cells (BRL) were cultivated on bronectin-coated 2D at glass surfaces or in different bronectin-coated 3D
scaffolds of different stiffness and architecture, xed and stained for F-actin (green) and nucleus (DAPI, blue). The rst column shows the maximum projection of a confocal image
stack with the DIC image and uorescence images merged. The second column gives the maximum projection of DIC only and the third column presents a 3D reconstruction. (A) A
BRL cell adherent on a 2D at glass surface. (B) A BRL cell growing in a segment of a stiff 3D scaffold made of PETTA. (C) BRL cells attached to a soft 3D scaffold made of Ormocomp
where the bending of the beams by the cells is clearly visible in the DIC image. The white arrows indicate the direction of view. (For interpretation of the references to color in this
gure legend, the reader is referred to the web version of this article.)

imaging of DAPI-stained interphase nuclei. A constant nucleus-tocell volume ratio (N/C ratio) is known to be important for maintaining cell integrity and tissue- and cell-specic functions,
whereas deviating N/C ratios have been observed in metastatic
cancer cells [46]. Mirroring the cell volume results, nuclear volumes
were increased in all broblast-like cell lines (BRL, B16, MEF)
residing in 3D scaffolds compared to a 2D surface (Fig. 6A, ANOVA,

*p < 0.05). Overall, nuclear volumes in 3D (~620e720 mm3) were

about two- to three-times higher than in 2D (~260 mm3) for all
broblast-like cell lines. In contrast, the nuclear volume of
epithelial cells was not signicantly different between 3D
(~420 mm3) and 2D (~380 mm3) culture conditions (Fig. 6A, ANOVA,
z, p > 0.05), again mirroring the cell volume results (Fig. 4D).
Calculating the N/C ratio under 2D and 3D culture conditions for

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

127

Fig. 3. Cell-matrix adhesions of cells on 2D substrates and in tailored 3D scaffolds. Buffalo rat liver cells (BRL) were cultivated on bronectin-coated 2D glass surfaces or bronectincoated 3D scaffolds of different mechanics and geometry, xed and stained for F-actin (green), paxillin-marked cell-matrix adhesions (red or white), and nucleus (DAPI, blue). The
rst column shows the maximum projection of a confocal image stack with the DIC image and uorescence images merged. The second column gives the maximum projection of
paxillin-stained cell-matrix adhesions only and the third column presents a 3D reconstruction of the cellular actin cytoskeleton. (A) A BRL cell adhering to a 2D glass surface. The
paxillin-containing cell-matrix adhesions are randomly distributed within the cells and most actin stress bers terminate in cell-matrix adhesion sites. (B) A BRL cell growing in a
stiff 3D scaffold made of PETTA. The paxillin staining is mainly localized along the beams of the scaffold. (C) BRL cells attached to a soft 3D scaffold made of Ormocomp. Paxillinpositive adhesion sites are predominantly located along the beams of the 3D structure. The white arrows indicate the direction of view. (For interpretation of the references to color
in this gure legend, the reader is referred to the web version of this article.)

broblast-like cells (BRL, B16, MEF) revealed no signicant changes

(Fig. 6B, ANOVA, z, p > 0.05). This is in agreement with the observation that both cell and nuclear volume of broblast-like cells
increase to a similar degree upon culturing in 3D environments

compared to 2D surfaces (compare 4D with Fig. 6A). The N/C ratio

was also maintained in epithelial cells (A549, NRK) for 3D and 2D
culture conditions (Fig. 6B, ANOVA, z, p > 0.05). This in turn is
consistent with the fact that these epithelial cells do not change

128

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

Fig. 4. The cell volume varies between 2D and 3D micro-environments in a cell-type dependent manner. (A) Maximum projection of a confocal image stacks of different cells types
(buffalo rat liver cells (BRL), spindle-shaped subpopulation of mouse melanoma cells (B16), mouse embryonic broblasts (MEF), human lung carcinoma cells (A549), normal rat
kidney cells (NRK)) adherent to stiff PETTA 3D scaffold. DIC image and uorescence images (F-actin green; nucleus DAPI, blue) are merged. (B) 3D reconstruction of the
according image in (A). The white asterisks indicate the angle of view between the maximum projection image and the 3D reconstruction. (C) 3D reconstruction of atomic force
microscopy (AFM) height images of cells on 2D glass surfaces in top view. White color in the look-up-table indicates a height of one mm or more. (D) The volume (mm3) of the
different cells types in the 3D structure and on a 2D surface was determined using the actin channel of the confocal image stacks (for 3D) or AFM measurements (for 2D). Fibroblastlike cells (BRL, B16, MEF) revealed a larger cell volume in 3D substrates compared to 2D surfaces. Epithelial cells (A549, NRK) showed no difference concerning their cell volume
between being cultured in 2D and 3D (ANOVA, *, p < 0.05; z, p > 0.05). The data were obtained from at least four independent experiments; the number of analyzed cells is indicated
in the according bars. The error bars are given as standard deviation of the mean. (For interpretation of the references to color in this gure legend, the reader is referred to the web
version of this article.)

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

Fig. 5. The volume of BRL cells is increased upon culture in 3D micro-environments.

The cell volume (in mm3) of buffalo rat liver cells (BRL) was determined using the
actin channel of confocal image stacks (for 3D scaffolds) or AFM measurements (for 2D
substrates). The cell volume of BRL cells was signicantly increased in stiff and soft 3D
micro-environments (dark gray and light gray bars) and in pure Matrigel (white bar),
in comparison to the 2D surface (black bar, ANOVA, *, p < 0.05). The data were obtained
from at least four independent experiments; the number of analyzed cells is indicated
in the according bars. The error bars are given as standard deviation of the mean.

their cell or nuclear volume during the transition from 2D to 3D

environments (compare 4D with Fig. 6A). However, we observed
differences in the proportion of nucleus to overall cell volume between the various cell types. While the nucleus in NRK cells constitutes 50% of the total cell volume, it takes up only 30e40% in the
other cell types (BRL, B16, MEF, A549; Fig. 6B). Overall, these results
indicate that changes in cell and nuclear volume in 3D vs. 2D environments is cell type-dependent, but that the N/C ratio is unaffected by the substrate dimension.
4. Discussion
The in vivo ECM is not only characterized by the composition of
numerous adherent and soluble molecules, but also by its 3D-geometry and mechanical properties. Here, we demonstrate that
DLW-fabricated 3D scaffolds can be used to systematically study
the impact of these various factors on cell-type specic behavior.
Fibroblast-like cell lines exhibit increased cell and nuclear volumes
in all investigated 3D scaffold types compared to 2D substrates,
independently of 3D substrate mechanics and architecture. In

129

contrast, epithelial cells preserve similar cell and nuclear volumes

in 2D and 3D environments. However, N/C ratios remain roughly
constant for all cell types and culture conditions. Cell-matrix adhesions forming in stiff and soft 3D micro-environments feature a
similar composition of prominent adhesion marker proteins as
adhesions in 2D environments.
Adherent cells require stable anchorage to their microenvironment and the generation of intracellular tension to divide
[47e50]. Mechanical matrix properties also have a signicant inuence on cell proliferation [51e53]. Additionally, is has been
demonstrated that the proliferation rate depends on matrix
composition and substrate dimensionality [16,22]. Since we found
similar cell proliferation rates of cells on 2D and in 3D microenvironments, we conclude that 3D DLW-structures provide sufcient support for maintaining cell growth and proliferation rates.
Cell-matrix adhesions are important cellular sensors of matrix
rigidity, as they mechanically link the ECM with the acto-myosin
system [9,10,18]. Compared to the 2D situation, relatively little is
known about the structure and function of cell-matrix adhesions
formed in 3D environments. Based on the markers investigated in
our study, we did not observe major differences in the composition
of cell-matrix adhesions of cells grown in 3D scaffolds in comparison to 2D substrates. Nevertheless, we noticed small morphological differences of cell-matrix adhesions between 2D and 3D culture
conditions. In 2D culture cell-matrix adhesions were usually elliptical, whereas in 3D culture adhesions were either highly elongated
when localized on features of the DLW structures. Our observations
are consistent with several studies that describe the general existence of cell-matrix adhesions in 3D culture substrates, but reports
vary depending on the adhesion markers and 3D matrix used. For
example, it has been shown that cell-derived matrices are more
effective in mediating broblast adhesion than 2D substrates or 3D
collagen gels, and that cells in soft 3D cell-derived matrices develop
so-called 3D cell-matrix adhesions, which differ in their molecular
composition from those formed on planar 2D substrates
[16,22e25]. Additionally, it was revealed that multiple cancer and
mesenchymal cell lines produce cell-matrix adhesions in 3D
collagen matrices depending on the cells' localization within the
matrix (close to the edges or to the center of a 3D collagen matrix)
[19e21,54]. We noticed in our study a slight reduction of FAK and
phospho-FAK staining intensity in both 3D scaffold types compared
to 2D substrates. Other studies showed increased FAK phosphorylation in dense and stiff collagen gels in comparison to softer gels in

Fig. 6. The nuclear volume increases in a cell-type dependent manner in 3D scaffolds while the nucleus-to-cell volume ratio is maintained. The cell volume and the nuclear
interphase volume (mm3) of different cells types (buffalo rat liver cells (BRL), spindle-shaped subpopulation of mouse melanoma cells (B16), mouse embryonic broblasts (MEF),
human lung carcinoma cells (A549), normal rat kidney cells (NRK)) on 2D glass surfaces and in stiff 3D PETTA scaffolds was determined. (A) The nuclear volume of cells residing in
3D substrates increased about two- to three-fold in broblast-like cells compared to the nuclear volume on 2D surfaces. Epithelial-like cells revealed no nuclear volume change
upon culturing in 3D structures. (B) The nucleus-to-cell volume ratio (N/C ratio) was maintained for all cell types (broblast-like vs. epithelial-like cells) under all culture conditions
(2D vs. 3D). While the nucleus occupied more than 50% of the cell volume in NRK cells, it was only about 20%e40% for BRL cells, B16 cells, MEF, and A549 cells (ANOVA, *, p < 0.05; z,
p > 0.05). The data were obtained from at least four independent experiments; the number of analyzed cells is indicated in the according bars. The error bars are given as standard
deviation of the mean.

130

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

epithelial cells [55] and reduced FAK phosphorylation in 3D

collagen gels in smooth muscle cells [56]. Nevertheless, a number of
studies are in apparent conict regarding the role of substrate
dimensionality and mechanics on the existence of cell-matrix
adhesion formation, composition, morphology, and function
[10,11,13,16,19e21].
Cell shape changes rely on the spatial and temporal coordination of biochemical and physical processes at the molecular, cellular
and tissue scale, and the homeostasis of cell volume is a major
challenge for eukaryotic cells and tissues [57e60]. Cell volume and
morphology may vary considerably under different environmental
conditions. For example, (i) cells regulate their volume in response
to changes in osmolarity of both extracellular and intracellular
environments [57,61]; (ii) cells show cell-cycle related volume
changes and in particular dividing cells dramatically change in
volume [59,60,62e64]; (iii) the volume of immune cells [65] or
metastasizing cells [13,14,64,66] changes when these cells squeeze
through narrow connements; and (iv) cell volume changes are
observed during epithelialemesenchymal transition in embryonic
development [67]. Changes in cell volume are also associated with a
broad spectrum of patho-physiological conditions. For cancer it is
known that trans-epithelial and cell volume regulatory ion transporters and channels are dys-regulated [58,61,64,66]. Using structurally dened DLW-fabricated scaffolds, we now demonstrate an
additional cell-type dependent effect on cell and nuclear volume
regulation in 2D and 3D environments. While broblast-like cell
lines exhibit increased cell and nuclear volumes in stiff 3D scaffolds
compared to 2D substrates, epithelial cells display similar volumes
in both (3D and 2D) environments. Volume measurements have
been mainly performed on cells cultured on planar and rigid (ECMcoated) substrates. In general, the cell volumes we obtained in our
study were in the range of volumes measured for various cells in 2D
studies (e.g., osteosarcoma cells: 1900 mm3 [68]; human osteoblasts: ~5000 mm3 [68]; human dermal broblast: 2400 mm3 [69];
rapidly proliferating mesenchymal stem cells: 1200e2100 mm3
[68]) and for further cell types [61,68e71]. However, so far only few
studies have analyzed single cell volumes in articial 3D environments [72,73]. For example, Farooque et al. investigated the volume
of human bone marrow stromal cells (hBMSCs) cultured in 3D gels
made of collagen or poly(ethylene glycol)tetramethacrylate and in
3D scaffolds of different geometries fabricated from poly(D,L-lactic
acid) or poly-(-caprolactone) [72]. Strikingly, hBMSCs volumes
obtained in this study were on average two to ve times (poly(D,Llactic acid scaffolds)) or even up to 20-times (free-form fabricated
poly--caprolactone scaffolds) larger than the cell volumes we
determined in our study; despite a similar approach for the
calculation of single cell volumes. These results point to large celltype specic differences in 3D volume regulation between stem
cells and broblast-like and epithelial cells. Furthermore, Farooque
et al. also obtained large cell-to-cell deviations in hBMSC volume on
the same 3D substrate [72], while we observed a more narrow
spread of cellular volume on a given substrate. Further differences
in 3D cell volume regulation could arise from the fact that different
materials and architectures were used compared to our study,
resulting in different chemical and physical signals received by the
cells. Additionally, Farooque et al. used culture times of up to
24 hours, while we analyzed cell volumes after three hours of
culture. In our 3D scaffold systems cells were well-spread after
three hours and we did not extend culture times further to avoid a
potential removal of the bronectin coating from the scaffold.
Nevertheless, we performed control experiments for broblasts
cultured in 3D scaffolds for 22 hours and observed that our 3D
scaffolds are suitable for longer cell culture times. However, we
detected not signicant differences in cell volume after 22 hours
compare to three hour of culture duration (data not shown).

Accurately measuring the volume of cells in complex environments presents an ongoing challenge in many areas of experimental and diagnostic science. Many different techniques have
been applied to provide estimates of single cell volumes, including
electrophysiological methods, interferometry, optical sectioning by
confocal microscopy, light scattering detection, quantitative uorescence or phase contrast microscopy, and AFM measurements
[68]. Unfortunately, often only relative changes in volume are
given, rather than absolute values which complicate the comparison between different studies [57,74,75]. We also realized that most
studies only investigate the volume of single cells that are cultured
either at 2D conditions or in 3D environments, while changes in cell
and nuclear volume during the transition from 2D to 3D cell culture
environments have rarely been reported. Our results therefore
identify an important but often disregarded aspect of cell type
specic behavior that is subject to cell volume regulation. Scaffolds
produced by DLW can be used to investigate such cell type-specic
differences in volume regulation in dened 3D environments.
Generally, cell volume and pressure regulation incorporates the
spatio-temporal synchronized function of water permeation, ion
channels, and transporters [57e59,64,74]. Although the precise
mechanisms have yet to be elucidated, early signals of volume
perturbation have also been demonstrated to involve the clustering
of integrin adhesion receptors into focal adhesion complexes and
the initiation of downstream signaling events such as the activation
of e.g., Rho family GTPases, phospholipases, lipid kinases, and
tyrosine- and serine/threonine protein kinases [58,64,76]. Adhesion receptor-mediated processes may therefore play a role in cell
volume regulation. However, we did not observe a signicant
change in the molecular composition of cell-matrix adhesions with
the markers tested here under the different 2D and 3D culture
conditions. Nevertheless, cell-matrix adhesions sites contain a large
number of adhesion proteins, and some variations in the composition of adhesion sites between 2D and 3D environments might be
expected inuencing cell volume regulation [10,11,16,19e21,23].
We also assessed the cell volume of BRL cells in pure Matrigel
lacking a DLW-fabricated 3D scaffold. Here, cells were signicantly
larger than on 2D substrates, similar as in both 3D scaffold types
(stiff and soft). BRL cells in pure Matrigel were mainly spherical
with some elongated protrusions, resulting sometimes in an
amoeboid to star-like cell morphology. An amoeboid cell
morphology and formation of cell-matrix adhesions were reported
in previous studies on different broblast cells cultivated in other
3D gel-like substrates [9e11,13,16,19,22]. In contrast, we did not
detect distinct paxillin-positive cell matrix adhesions in BRL cells
cultured in Matrigel. However, this is likely due to the short cultivation time of three hours because we observed dot-like paxillinpositive cell-matrix adhesions of BRL cells after Matrigel culture for
22 hours (data not shown).
Our study also revealed that both nuclear and total cell volume
of broblast-like cells increased in 3D vs. 2D micro-environments.
Interestingly, the N/C ratio remained unchanged for each cell line
under all culture conditions. A constant N/C ratio appears important for maintaining cell integrity and normal function. For
example, a decreased N/C ratio is an indication for metastatic cells
and is often used as a metric in cancer detection [46]. In breast
cancer cells the nuclear volume increases from normal to metastatic cell stages by a factor of about 1.6 for the cell volume and a
factor of two for the nuclear volume. Since we and others observed
that increased nuclear volume correlates with an increase in cell
volume [77e79], the question arises how total cell and nuclear
volumes could be co-controlled. Potential scenarios include water
inux through nuclear pores, signaling downstream of ECM and
growth factor receptors, and mechanical coupling between the
nucleus and the cytoskeleton [80e82]. For instance, nuclear

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

volume could be regulated by mechanical tension, organization of

the actin network or other cytoskeleton elements, and by how the
nucleus is connected to cytoskeletal laments. In fact it has been
demonstrated that reducing mechanical tension in the cytoskeleton leads to a decrease in nuclear size of endothelial cells [83].
In contrast to broblast-like cells, epithelial cells displayed
similar cell and nuclear volume in 2D and 3D micro-environments.
The constant cell volume of epithelial cells in tailored 3D structures
and on 2D substrate might be linked to the in vivo environment of
these cells: the ECM of different tissues is known to vary considerably in structure and composition, as does the morphology and
size of the respective matrix-embedded cells. For instance, polarized epithelial cells reside in vivo on a planar basal lamina containing a large amount of densely packed ECM proteins. In contrast,
broblasts reside in a 3D environment of loosely organized connective tissue.
5. Conclusion
Dimensionality-dependent effects on total cell and nuclear
volume expand our understanding of cell behavior in 3D environments. Cell and nuclear volume regulation seem to reect the
tissue-specic arrangement of cells in vivo such as loose connective
tissues (broblasts) or connement into at layers (epithelial cells).
In future, it will be of interest to investigate volume-sensitive
signaling pathways to elucidate in more detail the inuence of
substrate architecture on cell volume regulation processes. Potentially interesting pathways include second messenger release (e.g.,
Ca2 signaling), phosphorylation levels of proteins involved in
volume regulation (e.g., caveolin-1, cortactin, FAK), signaling via
integrins and receptor tyrosine kinases, and cytoskeletal reorganization [58,59,64,76]. Of interest will also be experiments studying
cell growth and cell volume regulation in response to different
ECM-ligand composition and the ligand's density. This will allow us
to progress beyond investigating only effects of altered dimensionality and mechanics on cell behavior. Overall, this will lead to a
more detailed understanding of cell behavior in engineered 3D
micro-environments that more accurately emulate in vivo situations. Generally, studying cells in 3D environments will be helpful
elucidating physiological and pathological aspects such as the immune response and cancer progression.
Acknowledgments
We thank Tanja Landmann for her help with the cell culture
routine and for her assistance performing cell experiments. C.M.
Franz acknowledges nancial support from the Deutsche Forschungsgemeinschaft (DFG) and the State of Baden-Wrttemberg
through the DFG-Center for Functional Nanostructures (CFN)
within subprojects E2.4.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.biomaterials.2015.08.016.
References
[1] W.W. Ahmed, T. Wolfram, A.M. Goldyn, K. Bruellhoff, B.A. Rioja, M. Moller, et
al., Myoblast morphology and organization on biochemically micro-patterned
hydrogel coatings under cyclic mechanical strain, Biomaterials 31 (2010)
250e258.
[2] M. Thery, Micropatterning as a tool to decipher cell morphogenesis and
functions, J. Cell Sci. 123 (2010) 4201e4213.
[3] M. Thery, A. Pepin, E. Dressaire, Y. Chen, M. Bornens, Cell distribution of stress
bres in response to the geometry of the adhesive environment, Cell Motil.
Cytoskelet. 63 (2006) 341e355.

131

[4] A.C. von Philipsborn, S. Lang, Z. Jiang, F. Bonhoeffer, M. Bastmeyer, Substratebound protein gradients for cell culture fabricated by microuidic networks
and microcontact printing, Sci. STKE 2007 (2007) pl6.
[5] M. Ochsner, M. Textor, V. Vogel, M.L. Smith, Dimensionality controls cytoskeleton assembly and metabolism of broblast cells in response to rigidity
and shape, PLoS One 5 (2010) e9445.
[6] M. Andreasson-Ochsner, G. Romano, M. Hakanson, M.L. Smith, D.E. Leckband,
M. Textor, et al., Single cell 3-D platform to study ligand mobility in cell-cell
contact, Lab Chip 11 (2011) 2876e2883.
[7] M. Ochsner, M.R. Dusseiller, H.M. Grandin, S. Luna-Morris, M. Textor, V. Vogel,
et al., Micro-well arrays for 3D shape control and high resolution analysis of
single cells, Lab Chip 7 (2007) 1074e1077.
[8] Y. Guan, W. Kisaalita, Cell adhesion and locomotion on microwell-structured
glass substrates, Colloids Surf. B Biointerfaces 84 (2011) 35e43.
[9] A.D. Doyle, R.J. Petrie, M.L. Kutys, K.M. Yamada, Dimensions in cell migration,
Curr. Opin. Cell Biol. 25 (2013) 642e649.
[10] B.M. Baker, C.S. Chen, Deconstructing the third dimension: how 3D culture
microenvironments alter cellular cues, J. Cell Sci. 125 (2012) 3015e3024.
[11] K.E. Kubow, S.K. Conrad, A.R. Horwitz, Matrix microarchitecture and myosin II
determine adhesion in 3D matrices, Curr. Biol. 23 (2013) 1607e1619.
[12] A.M. Greiner, B. Richter, M. Bastmeyer, Micro-engineered 3D scaffolds for cell
culture studies, Macromol. Biosci. 12 (2012) 1301e1314.
[13] A. Pathak, S. Kumar, Biophysical regulation of tumor cell invasion: moving
beyond matrix stiffness, Integr. Biol. (Camb.) 3 (2011) 267e278.
[14] S. Kumar, V.M. Weaver, Mechanics, malignancy, and metastasis: the force
journey of a tumor cell, Cancer Metast. Rev. 28 (2009) 113e127.
[15] R. Pankov, Y. Endo, S. Even-Ram, M. Araki, K. Clark, E. Cukierman, et al., A Rac
switch regulates random versus directionally persistent cell migration, J. Cell
Biol. 170 (2005) 793e802.
[16] E. Cukierman, R. Pankov, D.R. Stevens, K.M. Yamada, Taking cell-matrix adhesions to the third dimension, Science 294 (2001) 1708e1712.
[17] Y. Mao, J.E. Schwarzbauer, Stimulatory effects of a three-dimensional microenvironment on cell-mediated bronectin brillogenesis, J. Cell Sci. 118
(2005) 4427e4436.
[18] H.B. Schiller, R. Fassler, Mechanosensitivity and compositional dynamics of
cell-matrix adhesions, EMBO Rep. 14 (2013) 509e519.
[19] K.E. Kubow, A.R. Horwitz, Reducing background uorescence reveals adhesions in 3D matrices, Nat. Cell Biol. 13 (2011) 3e5 author reply-7.
[20] S.I. Fraley, Y. Feng, R. Krishnamurthy, D.H. Kim, A. Celedon, G.D. Longmore, et
al., A distinctive role for focal adhesion proteins in three-dimensional cell
motility, Nat. Cell Biol. 12 (2010) 598e604.
[21] S.I. Fraley, Y.F. Feng, D. Wirtz, G.D. Longmore, Reply: reducing background
uorescence reveals adhesions in 3D matrices, Nat. Cell Biol. 13 (2011)
5eU254.
[22] E. Cukierman, R. Pankov, K.M. Yamada, Cell interactions with threedimensional matrices, Curr. Opin. Cell Biol. 14 (2002) 633e639.
[23] J.S. Harunaga, K.M. Yamada, Cell-matrix adhesions in 3D, Matrix Biol. 30
(2011) 363e368.
[24] A.D. Doyle, M.L. Kutys, M.A. Conti, K. Matsumoto, R.S. Adelstein, K.M. Yamada,
Micro-environmental control of cell migrationemyosin IIA is required for
efcient migration in brillar environments through control of cell adhesion
dynamics, J. Cell Sci. 125 (2012) 2244e2256.
[25] K.M. Hakkinen, J.S. Harunaga, A.D. Doyle, K.M. Yamada, Direct comparisons of
the morphology, migration, cell adhesions, and actin cytoskeleton of broblasts in four different three-dimensional extracellular matrices, Tissue Eng.
Part A 17 (2011) 713e724.
[26] W.M. Petroll, L. Ma, Direct, dynamic assessment of cell-matrix interactions
inside brillar collagen lattices, Cell Motil. Cytoskelet. 55 (2003) 254e264.
[27] J.A. Pedersen, M.A. Swartz, Mechanobiology in the third dimension, Ann.
Biomed. Eng. 33 (2005) 1469e1490.
[28] P. Tayalia, E. Mazur, D.J. Mooney, Controlled architectural and chemotactic
studies of 3D cell migration, Biomaterials 32 (2011) 2634e2641.
[29] Y. Yang, F.M. Rossi, E.E. Putnins, Ex vivo expansion of rat bone marrow
mesenchymal stromal cells on microcarrier beads in spin culture, Biomaterials
28 (2007) 3110e3120.
[30] P. Soman, J.A. Kelber, J.W. Lee, T.N. Wright, K.S. Vecchio, R.L. Klemke, et al.,
Cancer cell migration within 3D layer-by-layer microfabricated photocrosslinked PEG scaffolds with tunable stiffness, Biomaterials 33 (2012)
7064e7070.
[31] B.K. Mann, A.S. Gobin, A.T. Tsai, R.H. Schmedlen, J.L. West, Smooth muscle cell
growth in photopolymerized hydrogels with cell adhesive and proteolytically
degradable domains: synthetic ECM analogs for tissue engineering, Biomaterials 22 (2001) 3045e3051.
[32] A.M. Greiner, M. Jackel, A.C. Scheiwe, D.R. Stamow, T.J. Autenrieth, J. Lahann,
et al., Multifunctional polymer scaffolds with adjustable pore size and chemoattractant gradients for studying cell matrix invasion, Biomaterials 35
(2014) 611e619.
[33] B. Richter, T. Pauloehrl, J. Kaschke, D. Fichtner, J. Fischer, A.M. Greiner, et al.,
Three-dimensional microscaffolds exhibiting spatially resolved surface
chemistry, Adv. Mater. 25 (2013) 6117e6122.
[34] A.S. Quick, J. Fischer, B. Richter, T. Pauloehrl, V. Trouillet, M. Wegener, et al.,
Preparation of reactive three-dimensional microstructures via direct laser
writing and thiol-ene chemistry, Macromol. Rapid Commun. 34 (2013)
335e340.
[35] F. Klein, T. Striebel, J. Fischer, Z. Jiang, C.M. Franz, G. von Freymann, et al.,

132

[36]

[37]

[38]

[39]

[40]
[41]
[42]

[43]

[44]

[45]
[46]

[47]
[48]
[49]

[50]

[51]
[52]
[53]

[54]

[55]

[56]

[57]

[58]

A.M. Greiner et al. / Biomaterials 69 (2015) 121e132

Elastic fully three-dimensional microstructure scaffolds for cell force measurements, Adv. Mater. 22 (2010) 868e871.
F. Klein, B. Richter, T. Striebel, C.M. Franz, G. von Freymann, M. Wegener, et al.,
Two-component polymer scaffolds for controlled three-dimensional cell
culture, Adv. Mater. 23 (2011) 1341e1345.
A. Ovsianikov, S. Schlie, A. Ngezahayo, A. Haverich, B.N. Chichkov, Two-photon
polymerization technique for microfabrication of CAD-designed 3D scaffolds
from commercially available photosensitive materials, J. Tissue Eng. Regen.
Med. 1 (2007) 443e449.
T. Weiss, R. Schade, T. Laube, A. Berg, G. Hildebrand, R. Wyrwa, et al., Twophoton polymerization of biocompatible photopolymers for microstructured
3D biointerfaces, Adv. Eng. Mater. 13 (2011) B264eB273.
K. Liefeith, T. Weiss, G. Hildebrand, R. Schade, Two-Photon polymerization for
microfabrication of three-dimensional scaffolds for tissue engineering application, Eng. Life Sci. 9 (2009) 384e390.
J. Fischer, M. Wegener, Three-dimensional optical laser lithography beyond
the diffraction limit, Laser Photonics Rev. 7 (2013) 22e44.
S. Maruo, O. Nakamura, S. Kawata, Three-dimensional microfabrication with
two-photon-absorbed photopolymerization, Opt. Lett. 22 (1997) 132e134.
A. Ovsianikov, A. Deiwick, S. Van Vlierberghe, P. Dubruel, L. Moller, G. Drager,
et al., Laser fabrication of three-dimensional CAD scaffolds from photosensitive gelatin for applications in tissue engineering, Biomacromolecules 12
(2011) 851e858.
R. Zanella, G. Zanghirati, R. Cavicchioli, L. Zanni, P. Boccacci, M. Bertero, et al.,
Towards real-time image deconvolution: application to confocal and STED
microscopy, Sci. Rep. 3 (2013) 2523.
Y. Zhang, B.J. Matuszewski, L.K. Shark, C.J. Moore, Medical image segmentation
using new hybrid level-set method, in: Medivis 2008: Fifth International
Conference Biomedical Visualization e Information Visualization in Medical
and Biomedical Informatics, Proceedings, 2008, pp. 71e76.
^u, M.-F. Devaux, Computation of Minkowski measures on 2D
D. Legland, K. Kie
and 3D binary images, Image Anal. Stereol. 26 (2007) 83e92.
V. Nandakumar, L. Kelbauskas, K.F. Hernandez, K.M. Lintecum, P. Senechal,
K.J. Bussey, et al., Isotropic 3D nuclear morphometry of normal, brocystic and
malignant breast epithelial cells reveals new structural alterations, PLoS One 7
(2012) e29230.
D.E. Ingber, Tensegrity-based mechanosensing from macro to micro, Prog.
Biophys. Mol. Biol. 97 (2008) 163e179.
I.B. Bischofs, S.A. Safran, U.S. Schwarz, Elastic interactions of active cells with
soft materials, Phys. Rev. E Stat. Nonlin. Soft Matter Phys. 69 (2004) 021911.
P. Moreo, J.M. Garcia-Aznar, M. Doblare, Modeling mechanosensing and its
effect on the migration and proliferation of adherent cells, Acta Biomater. 4
(2008) 613e621.
S.R. Peyton, C.M. Ghajar, C.B. Khatiwala, A.J. Putnam, The emergence of ECM
mechanics and cytoskeletal tension as important regulators of cell function,
Cell Biochem. Biophys. 47 (2007) 300e320.
P.C. Georges, P.A. Janmey, Cell type-specic response to growth on soft materials, J. Appl. Physiol. 98 (2005) 1547e1553.
M.J. Paszek, V.M. Weaver, The tension mounts: mechanics meets morphogenesis and malignancy, J. Mammary Gland Biol. Neoplas. 9 (2004) 325e342.
M.J. Paszek, N. Zahir, K.R. Johnson, J.N. Lakins, G.I. Rozenberg, A. Gefen, et al.,
Tensional homeostasis and the malignant phenotype, Cancer Cell. 8 (2005)
241e254.
S. Geraldo, A. Simon, N. Elkhatib, D. Louvard, L. Fetler, D.M. Vignjevic, Do
cancer cells have distinct adhesions in 3D collagen matrices and in vivo? Eur.
J. Cell Biol. 91 (2012) 930e937.
M.A. Wozniak, R. Desai, P.A. Solski, C.J. Der, P.J. Keely, ROCK-generated
contractility regulates breast epithelial cell differentiation in response to the
physical properties of a three-dimensional collagen matrix, J. Cell Biol. 163
(2003) 583e595.
I. Ishii, A. Tomizawa, H. Kawachi, T. Suzuki, A. Kotani, I. Koshushi, et al., Histological and functional analysis of vascular smooth muscle cells in a novel
culture system with honeycomb-like structure, Atherosclerosis 158 (2001)
377e384.
J.B. Holm, R. Grygorczyk, I.H. Lambert, Volume-sensitive release of organic
osmolytes in the human lung epithelial cell line A549: role of the 5-lipoxygenase, Am. J. Physiol. Cell Physiol. 305 (2013) C48eC60.
S.F. Pedersen, A. Kapus, E.K. Hoffmann, Osmosensory mechanisms in cellular
and systemic volume regulation, J. Am. Soc. Nephrol. 22 (2011) 1587e1597.

[59] H. Jiang, S.X. Sun, Cellular pressure and volume regulation and implications
for cell mechanics, Biophys. J. 105 (2013) 609e619.
[60] F. Lang, M. Ritter, N. Gamper, S. Huber, S. Fillon, V. Tanneur, et al., Cell volume
in the regulation of cell proliferation and apoptotic cell death, Cell Physiol.
Biochem. 10 (2000) 417e428.
[61] L. Allansson, S. Khatibi, T. Gustavsson, F. Blomstrand, T. Olsson, E. Hansson,
Single-cell volume estimation by three-dimensional wide-eld microscopy
applied to astroglial primary cultures, J. Neurosci. Methods 93 (1999) 1e11.
[62] M.P. Stewart, J. Helenius, Y. Toyoda, S.P. Ramanathan, D.J. Muller, A.A. Hyman,
Hydrostatic pressure and the actomyosin cortex drive mitotic cell rounding,
Nature 469 (2011) 226e230.
[63] P. Doroshenko, V. Sabanov, N. Doroshenko, Cell cycle-related changes in
regulatory volume decrease and volume-sensitive chloride conductance in
mouse broblasts, J. Cell Physiol. 187 (2001) 65e72.
[64] S.F. Pedersen, E.K. Hoffmann, I. Novak, Cell volume regulation in epithelial
physiology and cancer, Front. Physiol. 4 (2013) 233.
[65] T. Jin, X. Xu, D. Hereld, Chemotaxis, chemokine receptors and human disease,
Cytokine 44 (2008) 1e8.
[66] K.M. Stroka, H. Jiang, S.H. Chen, Z. Tong, D. Wirtz, S.X. Sun, et al., Water
permeation drives tumor cell migration in conned microenvironments, Cell
157 (2014) 611e623.
[67] S. Lamouille, J. Xu, R. Derynck, Molecular mechanisms of epithelialmesenchymal transition, Nat. Rev. Mol. Cell Biol. 15 (2014) 178e196.
[68] D. Docheva, D. Padula, C. Popov, W. Mutschler, H. Clausen-Schaumann,
M. Schieker, Researching into the cellular shape, volume and elasticity of
mesenchymal stem cells, osteoblasts and osteosarcoma cells by atomic force
microscopy, J. Cell Mol. Med. 12 (2008) 537e552.
[69] T. Imaizumi, F. Jean-Louis, M.L. Dubertret, C. Bailly, L. Cicurel, J.P. PetchotBacque, et al., Effect of human basic broblast growth factor on broblast
proliferation, cell volume, collagen lattice contraction: in comparison with
acidic type, J. Dermatol. Sci. 11 (1996) 134e141.
[70] E.A. Peeters, C.V. Bouten, C.W. Oomens, D.L. Bader, L.H. Snoeckx, F.P. Baaijens,
Anisotropic, three-dimensional deformation of single attached cells under
compression, Ann. Biomed. Eng. 32 (2004) 1443e1452.
[71] F. Lang, M. Ritter, E. Woll, H. Weiss, D. Haussinger, J. Hoacher, et al., Altered
cell volume regulation in ras oncogene expressing NIH broblasts, Pugers
Arch. 420 (1992) 424e427.
[72] T.M. Farooque, C.H. Camp Jr., C.K. Tison, G. Kumar, S.H. Parekh, C.G. Simon Jr.,
Measuring stem cell dimensionality in tissue scaffolds, Biomaterials 35 (2014)
2558e2567.
[73] L. Saias, J. Swoger, A. D'Angelo, P. Hayes, J. Colombelli, J. Sharpe, et al.,
Decrease in cell volume generates contractile forces driving dorsal closure,
Dev. Cell 33 (2015) 611e621.
[74] S.A. Harron, C.M. Clarke, C.L. Jones, D. Babin-Muise, E.A. Cowley, Volume
regulation in the human airway epithelial cell line Calu-3, Can. Can. J. Physiol.
Pharmacol. 87 (2009) 337e346.
[75] N. Iida-Tanaka, I. Namekata, M. Tamura, Y. Kawamata, T. Kawanishi,
H. Tanaka, Membrane-labeled MDCK cells and confocal microscopy for the
analyses of cellular volume and morphology, Biol. Pharm. Bull. 31 (2008)
731e734.
[76] S.F. Pedersen, E.K. Hoffmann, J.W. Mills, The cytoskeleton and cell volume
regulation, Comp. Biochem. Physiol. A Mol. Integr. Physiol. 130 (2001)
385e399.
[77] P. Roca-Cusachs, J. Alcaraz, R. Sunyer, J. Samitier, R. Farre, D. Navajas, Micropatterning of single endothelial cell shape reveals a tight coupling between
nuclear volume in G1 and proliferation, Biophys. J. 94 (2008) 4984e4995.
[78] J.G. Umen, The elusive sizer, Curr. Opin. Cell Biol. 17 (2005) 435e441.
[79] J.A. Swanson, M. Lee, P.E. Knapp, Cellular dimensions affecting the nucleocytoplasmic volume ratio, J. Cell Biol. 115 (1991) 941e948.
[80] J.A. Mellad, D.T. Warren, C.M. Shanahan, Nesprins LINC the nucleus and
cytoskeleton, Curr. Opin. Cell Biol. 23 (2011) 47e54.
[81] K.N. Dahl, A.J. Ribeiro, J. Lammerding, Nuclear shape, mechanics, and
mechanotransduction, Circ. Res. 102 (2008) 1307e1318.
[82] T. Shimi, V. Butin-Israeli, R.D. Goldman, The functions of the nuclear envelope
in mediating the molecular crosstalk between the nucleus and the cytoplasm,
Curr. Opin. Cell Biol. 24 (2012) 71e78.
[83] J.R. Sims, S. Karp, D.E. Ingber, Altering the cellular mechanical force balance
results in integrated changes in cell, cytoskeletal and nuclear shape, J. Cell Sci.
103 (Pt 4) (1992) 1215e1222.