Showcasing research from Department of Chemical As featured in:

Engineering, National Taiwan University and Institute of

Nanoengineering and Microsystems, National Tsing Hua
University, Taiwan

Using cell structures to develop functional nanomaterials

and nanostructures – case studies of actin filaments and
microtubules

Actins and microtubules are utilized as building blocks to create

functional nanomaterials and nanostructures for nature-inspired
small-scale devices and systems.

See Chao-Min Cheng et al.,

Chem. Commun., 2014, 50, 4148.

www.rsc.org/chemcomm
Registered charity number: 207890
 ChemComm
View Article Online
FEATURE ARTICLE View Journal | View Issue

Using cell structures to develop functional

Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

nanomaterials and nanostructures – case

Cite this: Chem. Commun., 2014,
50, 4148 studies of actin filaments and microtubules
Kevin Chia-Wen Wu,†a Chung-Yao Yang†b and Chao-Min Cheng*b

This article is based on the continued development of biologically relevant elements (i.e., actin filaments
and microtubules in living cells) as building blocks to create functional nanomaterials and nanostructures
Received 1st January 2014, that can then be used to manufacture nature-inspired small-scale devices or systems. Here, we summarize
Accepted 21st January 2014 current progress in the field and focus specifically on processes characterized by (1) robustness and ease
DOI: 10.1039/c4cc00005f of use, (2) inexpensiveness, and (3) potential expandability to mass production. This article, we believe,
will provide scientists and engineers with a more comprehensive understanding of how to mine
www.rsc.org/chemcomm biological materials and natural design features to construct functional materials and devices.

Introduction functional properties of the ‘‘marriage’’, using either micro-

tubules or actin filaments.
Humankind often derives guidance and inspiration from nature, The design of functional organic nanostructures using bio-
a wellspring for elegant, complex, integrated, and optimized templating is still an almost unexplored field. In this article, we
structural solutions that enable organisms to accomplish complex attempt to summarize current progress in the field and focus
functions. Unfortunately, many scientists approach design specifically on processes characterized by (1) robustness and
without any explicit reference to nature, as direct natural analogs ease of use, (2) inexpensiveness, and (3) potential expandability
do not exist for many associated technological applications. to mass production, through the novel use of actin filaments
In recent years however, there has been increasing interest in and microtubules.
borrowing design concepts or materials from nature to create The cytoskeleton is a filamentous network of F-actin, micro-
small-scale functional materials, structures, and systems. tubules, and intermediate filaments composed of one of three
Nanomaterials such as inorganic nanowires, metal-based nano- chemically distinct subunits: actin, tubulin, or one of several
wires or nanoparticles (NPs), quantum dots (QDs), or carbon groups of intermediate filament proteins. The complex dynamics
nanotubes demonstrate promising optical, electronic, and of actin filaments and other cytoskeletal elements play an impor-
catalytic properties. Biomolecules, such as DNA, enzymes, or tant role in multiple cellular behaviors such as cell division,
antibodies, contain dimensional similarities with these nano- motility, and determination of cell shape.2 Cytoskeletal filamentous
materials and this suggests that the combination/integration of (F-) actin, a double-stranded helical filament made of globular actin
biomolecules with nanomaterials may develop hybrid systems (G-actin), is, essentially, a semiflexible polymer and highly charged
that leverage collective properties.1 Considerable previous polyelectrolyte with a diameter of B8 nm and a persistence length
accomplishments, over recent decades, have advanced the develop- of B10 mm.3 Examination via X-ray diffraction and small-angle
ment of biomolecule-conjugated nanomaterials or nanostructures X-ray scattering could indicate the presence of minute structural
and their applications for nanoscale machinery, sensing, logic elements such as filaments, bundles, and networks at the mole-
operations, and nanodevices.1 However, little of this work has cular level that could provide insight into the interactive process
been directed toward developing a system in which both bio- of self-assembly of biological polyelectrolytes like actin.4
and nano-sourced building block contributors have mutually Actin, an intracellular structural protein, plays a crucial role
benefited and, in paired combination, improved the final in a wide range of cellular behaviors, including the migration of
eukaryotic cells, their mitosis, and their shape (mechanical
integrity).5 Actin stress fibers have also been known to actively
a
Department of Chemical Engineering, National Taiwan University, Taipei 10617,
remodel in response to cyclic strains, as demonstrated in cultured
Taiwan
b
Institute of Nanoengineering and Microsystems, National Tsing Hua University,
bovine aortic endothelial cells.6 In the cyclic strain research that
Hsinchu 30013, Taiwan. E-mail: [email protected] demonstrated this remodelling, bovine aortic endothelial cells were
† These authors contributed equally. seeded on fibronectin-coated silicone membranes and subjected

4148 | Chem. Commun., 2014, 50, 4148--4157 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Feature Article ChemComm

to either uniaxial or biaxial sinusoidal stretching at a controlled cytoskeletal conformation changes in NIH-3T3 fibroblasts after
frequency. Both of these dynamic loading conditions caused an a single heat shock treatment (37 1C to 43 1C).10 In this work,
increase in actin stress fiber density, but the changes in fiber with respect to cytoskeletal organization, we found that the
orientation were distinct as the output responses in these two cytoskeleton approached the original polymerized organization
techniques diverged. Frequency imposed uniaxial stretching following heat shock, but did not completely return to the
induced stress fibers to orient themselves perpendicular to the original state. At high temperatures, the lipid bilayer of cell
axis of stretching, while biaxial stretching did not result in any membranes can dissociate. The resulting changes in the
specific fiber orientation.7 Such alignment of stress fibers in membrane potentials could induce many possible changes in
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

the direction of minimal substrate deformation has also been cellular function, resulting from the loss of proteins critical for
reported by Wang et al.8 maintaining bilayer integrity.11
Fig. 1 reveals the actin filaments of a living 3T3 fibroblast Microtubules (MTs) are cytoskeletal, self-assembling,
and a porcine kidney-1 epithelial cell via liquid-type atomic dynamic, tubular structures with nanometerically sized diameters
force microscopy (AFM). It is well documented that mammalian and average lengths of 25 mm.12 The cross-section of a MT will
cells subjected to cyclic mechanical strain transmit this show a ring with inner and outer diameters of approximately 16
mechanical force into intracellular signals and induce cellular and 24 nm, respectively. Microtubules have several functions,
responses, and may suppress apoptosis.9 These results indicate e.g., they form the cytoskeleton lattice in the cytoplasm of
that cell proliferation would be enhanced by cyclic mechanical eukaryotic cells, axonal MTs act as tracks for transporting
stimulation. Physiomechanical stress affects the cytoskeleton particles, and MTs generate forces for movement in flagella
and plays an important role in determining and maintaining and cilia. They also play an essential role in mitosis during cell
cellular function. As with the physiomechanical stress or stimu- division. Microtubule formation can be mimicked in vitro
lation, thermal change/stimulation is likely to impart significant through the addition of guanosine-5 0 -triphosphate (GTP) to
conformational changes to actin via the cytoskeleton. This is a solution of the a- and b-tubulin heterodimer.13 Clearly, com-
based on our own previous research in which we investigated bining biomolecules with nanomaterials represents an area of

Fig. 1 (a) Revealing actin filaments from a living 3T3 fibroblast using AFM. A living 3T3 fibroblast was seeded on a glass coverslip coated with type I
collagen, and mounted onto a Bio-Cell (JPK Instrument, Germany) temperature-controlled sample holder filled with DMEM containing 10% FBS as
culture media. A Nano Wizard II (JPK Instruments, Germany) atomic force microscope was used to scan the cell in the liquid under contact mode. Silicon
nitride cantilevers (CSC-38B, MikroMasch) with a nominal spring constant of 0.03 N m1 (calibrated using the thermal noise method) and a cone half
angle o151 were used for imaging. Prior to scanning, the scanning range of the AFM cantilever tip was positioned to ensure full coverage of the cell. The
cell was then scanned with the tip velocity controlled between 90–110 mm s1 and force applied between 0.5–1.0 nN. The whole scanning time was kept
under 15 minutes to minimize the changes in the cytoskeleton throughout the process. The image file was processed using JPK data analysis software
(JPK Instruments, Germany), where the deflection, height, and three-dimensional and cross-section profile images could be extracted. The location of
the actin filaments was previously identified by comparing the scanned image of an actin–RFP-3T3 cell and its respective fluorescent image, where
prominent actin filaments could be observed. The linear filamentous structures were shown to be actin filaments. (b) Revealing the cortical actin
filaments of a living PK-1 (Porcine Kidney-1) epithelial cell using AFM. Notice the actin filaments in this cell are shorter compared to those of the
fibroblasts, and are located mainly on the cell periphery, fitting the description of so-called ‘‘cortical actin.’’ (a) and (b) are experimental work contributed
by Hans Harn and Dr Yao-Hsien Wang of Dr Ming-Jer Tang’s lab, respectively.

This journal is © The Royal Society of Chemistry 2014 Chem. Commun., 2014, 50, 4148--4157 | 4149
 View Article Online

ChemComm Feature Article

considerable promise and an intriguing array of possibilities in

the area of materials science. As a result of great interest and
novel research, MTs have been used to initiate transport of
biomolecules on engineered surfaces and to study the biochemistry
of motor proteins.14 In this article, we wish to accomplish two
objectives: (1) outline current progress, and (2) suggest ways to
overcome any obstacles inherent in making functional nano-
materials, nanostructures, and (ultimately) nanosystems using
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

biologically relevant elements, i.e., actin filaments and MTs.

Our goals are based on research and significant interest generated
in multiple academic communities including nanomaterials
(materials, chemical science), nanotechnology (engineering,
applied physics) and biotechnology (cell biology, bioengineering).

Preparation of individual actin

filaments and microtubules in vitro
For the preparation of individual actin filaments in vitro, we
describe one of the commonly used procedures.15 F-actin in an
ATP-buffer16 or phosphate buffered saline solution (PBS, from
Fisher) was first prepared. The steps to perform this were as
follows: (1) 1 mL of ATP-buffer was combined with 2 mL pure
ATP (100 mM); (2) G-actin was re-suspended to obtain a
concentration of 0.4 mg mL1; (3) this solution was mixed
and incubated for one hour at 24 1C after adding 66 mL ATP
polymerization buffer (from Cytoskeleton, USA; No.: BSA02);17
and, (4) 2 mL Milli-Q water was added to the solution, which
was then kept in liquid nitrogen (70 1C) for one hour with the Fig. 2 (a) Fluorescence image of F-actin. F-actin filaments were stained
with 6 mM Alexa Fluor 488 phalloidin. The samples were examined under
diluted G-actin before incubating for one hour at 24 1C. Fig. 2a
an epi-fluorescent microscope (Zeiss Axiovert, Germany) equipped with a
contains images of the polymerized actin filaments, which were 63 high numerical aperture (NA = 1.4) oil immersion objective to image the
5–10 mm length. These were created with different concentra- actin filaments (scale bar = 1 mm). In (a), the image shows that the actin
tions and labeled with phalloidin (from Molecular Probes, USA; filaments have condensed into a parallel arrangement and formed an
No.: A12379). The samples were examined under a fluorescent ensemble of thick bundles. It is noted that dramatically increased fluores-
cence intensity is obtained from the multifilament bundles. (b) and (c) are
microscope (Zeiss Axiovert, Germany) using a 63 high numerical
atomic force microscopy images of the actin filaments; (b) the matrix of actin
aperture (NA = 1.4) oil immersion objective to image the G-actin filaments and (c) the morphology of a single actin filament with a length of
monomer aggregates and F-actin. We also imaged the filamentous 1.8 mm. (d) SEM image of brain microtubules assembled after a single
actin using atomic force microscopy (AFM) in an aqueous environ- polymerization cycle. (e) A microtubule-rich region in the sample showing
ment as shown in Fig. 2b. This displays not only the distribution of contaminant material. (f) AFM topography images of polymerized micro-
tubules adsorbed to silanized mica, where a few individual microtubules with
actin filaments in a network, but also a higher resolution portrayal
a well preserved structure are easily observed. (d–f) are from ref. 18d.
of individual actin filaments on the surface of the material. We
continued to probe the behaviour of F-actin by obtaining the
deflection-displacement curve via AFM for single actin filaments Fig. 2d displays SEM images of the polymerized MTs, which were
with a silicon nitride tip (the spring constant of a cantilever is 500 nm–1.5 mm in length. These were assembled after a single
0.2 N m1); the adhesive force of the silicon nitride tip on the single polymerization cycle. Fig. 2e and f show AFM images of
actin filament was between 0.15 and 2 nN (Fig. 2c). MTs adsorbed to mica functionalized with APTES, a silane with
For the preparation of individual MTs in vitro, many similar an amine group that is positively charged at BpH 7. Mica is
MT protein purification methods have been developed.18 In composed of multi-layers, which can be peeled to generate a new
this study, we describe one of the commonly used procedures. clean surface without any preliminary washing procedures
Tubulin was purified from porcine brains by two assembly– (unlike traditional glass). Microtubules on mica were observed
disassembly cycles and phosphocellulose chromatography, and in tapping mode at atmosphere in order to minimize
stored in liquid nitrogen at 6 mg mL1. Tubulin was polymerized de-adsorption. Fig. 2e displays an AFM topography of microtubule-
into MTs in BRB80 buffer (1 mM EGTA, 1 mM MgCl2, 80 mM rich region in the sample showing contaminant material. An AFM
PIPES–NaOH) containing 1 mM GTP and 1 mM MgSO4. topography image of polymerized MTs adsorbed to silanized
After incubating at 37 1C for 30 minutes, the MTs were stabilized mica, where a few individual MTs with a well preserved structure
and diluted 100 in BRB80 containing 20 mM toxal.18d are easily observed, is shown in Fig. 2f.

4150 | Chem. Commun., 2014, 50, 4148--4157 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Feature Article ChemComm

Various applications of using actin/ structure of actin-containing liposomes.20c,d They found that
microtubule cytoskeletons in vitro actin-containing liposomes could be extruded through poly-
carbonate membranes with pore sizes of 400–600 nm. At low
Actin has long been known for providing strong structural support actin concentration (i.e., 0.1 mg mL1), the morphology of
for cells and for acting as a linker for protein motor systems these liposomes is spherical in shape. As actin concentration
important in biological systems. To date, several studies have was increased, the morphology changed to a disk-like shape
reviewed the structure, mechanism, and mechanical strength (1 mg mL1) and then back to a spherical shape (5 mg mL1).20c
of self-assembling actin and actin’s interactions with other These results indicate that the formation of a bundle of actin
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

biomolecules.19 Beyond examination of these more fundamental filaments can affect the morphology of soft lipid bilayers and
studies, there is growing interest and research into the use of alter structural strength.
actin/MTs for a range of applications (Fig. 3) because actin By encapsulating actin into the aqueous core of liposomes,
exhibits several advantages: (1) excellent biocompatibility, (2) the circulatory half-life was significantly increased from 19 hours
robustness and ease of use, (3) inexpensiveness, (4) realizable for plain liposomes (without actin) to 35 hours for actin-containing
and precise control of self-assembled structures and, (5) liposomes.20c,d Findings suggest that extended circulatory per-
potential expandability to mass production. Herein, several sistence is desirable for drug delivery applications, since
potential applications of actin/MTs including drug delivery, it avoids repeated drug administration. Actin filaments and
patterning, and nanofabrication are introduced. bundles can also be modified with myosin II to transport
liposomes (as ‘‘nano-containers’’ to encapsulate drugs or other
Drug delivery molecules of interest) and single cells across an inorganic
Actin, the major component of the cytoskeleton, provides surface.21 Biotinylated liposomes or biotinylated E. coli cells
functional mechanical stiffness for cells. Inspired by its role engineered to express green-fluorescent protein (GFP) were attached
in cellular architecture, several research groups have used actin to F-actin or actin bundles labeled with rhodamine–phalloidin and
filaments to stabilize the structure of liposome-based drug biotin–phalloidin via a neutravidin linkage (Fig. 4a). Fig. 4b shows
nanocarriers, because plain liposomes generally decompose that the actin bundles loaded with liposomes were found to move
when they are exposed to fluid shear stresses.20 Cortese et al. across the flow cell surface, while the F-actin preparation was
compressed actin filaments inside the core of liposomes and able to transport liposomes. Compared to using actin filaments
found that the deformation of the liposome was dependent on for drug delivery/transport, the use of microtubules as vehicles
the length of the encapsulated actin filament.20a Miyata and for drug delivery is more common.
Hotani found that liposomes encapsulating G-actin exhibited a In recent years, intracellular transport studies have indicated
spherical shape but changed to dumbbell and disk-like shapes the capacity for well-organized bidirectional nanotransport of
after G-actin was polymerized inside the liposomes.20b Li and individual molecules or vesicles, a process that is supported by
Palmer also studied the effect of actin concentration on the MTs preorganized for cargo transport by kinesin and dynein.22

Fig. 3 Schematic demonstrating the production of functional materials, structures, and systems at the nanometer scale using actin and microtubule
cytoskeletons.

This journal is © The Royal Society of Chemistry 2014 Chem. Commun., 2014, 50, 4148--4157 | 4151
 View Article Online

ChemComm Feature Article

Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

Fig. 4 Potential drug delivery applications using either actin filaments or microtubules. (a) Schematic of biotinylated liposomes and E. coli cells that have
been bound to biotin-labeled actin bundles via a neutravidin linker. (b) Time series images of actin bundles (red; arrowheads) transporting liposomes
(green; arrows). (a) and (b) are from ref. 21. (c) The mobility of kinesin and a microtubule dissociation method enable orientation of a microtubule in an
array for directed transport of reactive molecules or drugs carried by kinesin or dynein; ref. 23. (d) Wild type tau (htau40) binds to the MTs creating
‘‘roadblocks’’, and hinders the motion of the beads. The tau protein attached on the microtubule surface may lead to a detachment of the kinesin
molecule. The motion may be continued when another kinesin attached to the same bead binds onto the microtubule surface; ref. 24d.

This can be reconstructed in vitro as a bead assay-based system created by tau protein molecules bound to the MTs could be
by fixing MTs on a glass substrate and allowing motors to carry examined.24 Htau40 binds to MT protofilaments, hindering
cargos toward the direction defined by the polarities of the kinesin-coated bead motion (Fig. 4d). The P301L mutation
MTs. Two important technical challenges to realize further reduces the MT-binding capability of the tau protein, preventing
organized molecular systems based on a bead assay-based the generation of ‘‘roadblocks’’ and allowing the unhindered
system are controlling the cargo transport direction by defining movement of kinesin along the MTs. Multiple kinesin molecules
the polarities of individual MTs and employing both kinesin access the MT surface, jointly contributing to the bead movement.
and dynein to actively and bidirectionally transport cargos for Therefore, if a kinesin molecule encountering a ‘‘roadblock’’
biochemical assays. Fig. 4c shows the bead assay-based system (i.e., a tau protein) detaches from the microtubule protofila-
using multiple motors reconstructed in a nanotrack array ment, another kinesin may continue the motion (Fig. 4d).
proposed by Fujimoto et al.23 This system can regulate the As a result, tau protein ‘‘roadblocks’’ produce an effect of
direction of molecular transport and the physical interaction of ‘‘speed bumps’’ for the kinesin molecules, causing an overall
target molecules by using MTs as vehicles. Molecular transport slowing down of the kinesin-coated beads. By combining tau
resulting from the movements of kinesin and dynein caused a proteins and MTs, we suggest that a drug can be delivered
collision of the GST and GSH sequences, which was revealed by to precise positions.
Q-dot colocalization.
Another potential approach for drug delivery is binding Nanopatterning
tau protein molecules functionally characterized for their effect Conventional patterning methods are ‘‘top-down’’ approaches
on MT kinesin-based transport. The effect of ‘‘roadblocks’’ that involve the use of photolithography or electron-beam

4152 | Chem. Commun., 2014, 50, 4148--4157 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Feature Article ChemComm

lithography and subsequent etching and polishing processes. step in investigating structural characterization via X-ray
These methods are very useful for constructing two-dimensional diffraction, nuclear magnetic resonance (NMR), and transmis-
(2D) structures with various patterns (e.g., linear patterns, circular sion electron microscopy (TEM). Elucidating intramolecular
patterns, or mixed patterns).25 Although current nanotechnologies interactions in the aligned (or crystallized) structures would
can create patterns as small as several nanometers, expensive provide structural and mechanistic insights.
equipment and a clean environment are necessary, and this limits Molecular structure holds a key to understanding nature’s
the applicability for semiconductor industries. In contrast to intricate design mechanisms and blueprints. If we can under-
top-down approaches, molecular self-assembly of peptides and stand these blueprints and basic materials, perhaps we can
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

proteins is a ‘‘bottom-up’’ approach that can easily generate 2D begin to mimic the elegance of nature by making beautiful
or 3D patterns of nanometric size. Actin filaments, one of the products more cost effectively and with less detrimental environ-
cellular fibrous proteins, consist of self-associated monomers mental consequences. A famous example of leveraging molecular
of actin. Because actin monomers exhibit guanidine and structure characterization is the discovery of DNA structure
adenosine triphosphate hydrolytic activity that affords confor- (amyloid b-peptide) by Fairlie, Craik, and Watson via NMR
mational self-assembly, control of actin self-assembly is possible spectroscopy.27 Their results may provide new insight to the
and key to understanding the relationships between protein structural properties of amyloid b-peptides, which are of great
assembly and a number of diseases. relevance to Alzhemier’s disease research.
There are two forms of actin: F-actin is in the form of We have attempted to control actin filament organization
polymerized fibers, and G-actin is globular in form. Previous through biologically-inspired intermediates, allowing us to create
studies have shown that G-actin could transform to F-actin a regular nanopattern with a large area.26e We first placed
when the concentration of salt in the solution increased. individual actin filaments on a pristine glass surface (i.e., a
Methods for the alignment (patterning) of F-actin have also hydrophilic surface) and introduced myosin-II to modify this
been widely studied,26 as shown in Fig. 5. The stretch frame hydrophilic surface. We then employed the inorganic salt
method in Fig. 5a can create a bundle of fibers oriented along crystallization approach to probe the response of two proteins,
the fiber axis (1D alignment), whereas the dropping and drying actin filaments and myosin-II, in order to analyze the resultant
methods in Fig. 5b and c can only result in all the fibers lying spatially localized patterns. Fig. 6a and b show fluorescence
along the same plane. Examining and identifying the orienta- images of actin filaments on a pristine glass surface and
tion of F-actin or other fibrous proteins and peptides was a first actin filaments on a glass surface treated with myosin-II
(non-covalent immobilization), respectively. We quantified the
behaviors of both by importing fluorescence images of actin
filaments labeled with phalloidin into ImageJ software,28 and
measuring the angle between a single filament and its corres-
ponding horizontal axis to determine the alignment of the
filaments on pristine glass, and glass treated with myosin-II.
This analysis indicated that the orientation of actin filaments
placed on both pristine and functionalized surfaces exhibited
orientations of 121 and 381, respectively. In order to gain
further understanding of filament orientation behavior for
both samples, we used our previously described salt crystal-
lization method.26b Fig. 6c–e show salt crystallization images of
actin filaments on a pristine glass surface, only myosin-II on a
glass surface, and actin filaments on a glass surface treated
with myosin-II (1 mg mL1), respectively. The orientation of
filaments induced by salt crystallization of only actin filaments
on a glass surface (Fig. 6c) indicated that salt crystallization
induced a more ordered orientation as reflected in the observa-
tion of filaments branched in perpendicular directions.
On the other hand, Ionov et al. attempted to fabricate
stimuli-responsive nanopatterned polymer brushes using MTs
as templates.29 In their study, they used atom transfer radical
polymerization to initiate thermoresponsive poly-(N-isopropyl-
acrylamide) brushes on MTs. Rhodamine-labeled MTs were
prepared by self-assembly of a,b-tubulin dimers, and then
adsorbed on DDS-coated glass surfaces by perfusing them in
Fig. 5 Methods for aligning fibrous proteins through dehydration.
A solution of fibrils was suspended between two sealed capillaries to
buffer solution through a narrow channel between two glass
generate a bundle of fibers (a) or suspended inside a capillary (b) or on a coverslips. The MTs formed chemical links between neighboring
mat or film (c) to generate fibers on the same plane; ref. 26d. amino groups after crosslinking by glutaraldehyde and lysine

This journal is © The Royal Society of Chemistry 2014 Chem. Commun., 2014, 50, 4148--4157 | 4153
 View Article Online

ChemComm Feature Article

Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

Fig. 6 Actin filament organization on pristine and functionalized glass surfaces. (a) Fluorescence image of actin filaments on a pristine glass surface. (b)
Fluorescence image of actin filaments on a glass surface modified with myosin-II. Scale bar = 1 mm. Optical images of the salt crystallization related to the
actin filament presence for (c) only actin filaments on a pristine glass surface, (d) only myosin-II on a glass surface, and (e) actin filaments on a glass
surface modified with myosin-II, respectively. Scale bar = 1 mm; ref. 26e. Morphology of microtubules at different stages of the modification procedure
via AFM: (f) after crosslinking with glutaraldehyde, (g) after immobilization of initiator, and (h) after grafting of poly-(N-isopropylacrylamide-fluorescein
o-acrylate); ref. 29.

residues (Fig. 6f). After treating with 2,2 0 -(ethylenedioxy)bis- peptides.30 In contrast, gradual removal of salts would allow
(ethylamine) and bromo-2-methylpropanoyl bromide, the mod- one to re-dissolve charged peptides and arrange them in an
ified MTs possessed an undeformed rod-like shape (Fig. 6g). ordered direction. The driving forces for such macroscopic
The synthesized polymer brushes were modified on MTs by adding orientation include hydrophobic/hydrophilic forces and hydro-
fluorescein o-acrylate for the polymerization of N-isopropyl- gen bonding. A surface micropatterning method has also been
acrylamide (Fig. 6h). reported to geometrically control the growth and alignment of
Recently, macroscale alignment has attracted significant actin.31 This study further demonstrates that actin-filament
scientific attention. A salting-out strategy has been reported orientation could determine the interactions between two
to create macroscale parallel assays of peptide nanotubes.2a actin filaments and the formation of actin bundles. Because
The addition of anions (e.g., SO42) is efficient for precipitating it is known that the spatial and temporal modulation of
positively charged peptides, and such salt-induced precipita- the environment exerts great influence on actin cytoskeleton
tion is a well-known technology for purification of charged architectures, causing different cell morphologies, a defined

4154 | Chem. Commun., 2014, 50, 4148--4157 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Feature Article ChemComm

geometrical boundary condition could help clarify the effects of self-assembled peptide (i.e., the Alzheimer’s-amyloid diphenyl-
spatial arrangements within actin-filament network architectures. alanine structural motif) to form nanotubes.37 Silver nanowires
were then generated inside the peptide nanotubes by impregna-
Nanofabrication tion and reduction of silver ions. An enzyme (i.e., proteinase K)
Molecular self-assembly has been considered a ‘‘bottom-up’’ was then applied to degrade the diphenylalanine peptides,
approach to efficiently build up nanomaterials. DNA and self- resulting in discrete and enzymatically stable nanowires with a
assembled peptides have been widely used as templates for long, persistent length. In this section, we will mainly focus on
metal deposition. Metal nanowires such as silver,32 gold,33 the use of actin filaments to develop functional nanowires
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

platinum,34 palladium,35 and copper36 have been successfully because there is less research regarding the use of MTs.
synthesized using DNA as a template. Cationic metal ions can G-actin has also been used as a molecular building block for
effectively interact with the anionic phosphoric groups of DNA, creating metallic nanowires.38 Although G-actin has a globular
and subsequent reduction methods including UV irradiation, structure, it undergoes polymerization when ATP, Mg2+, and K+
thermal treatment, or chemical reducing agents can convert are present. This ATP-driven polymerization converts G-actin
metal ions into metal along the DNA template, resulting monomers to F-actin filaments, which is an important process
in metal nanowires of uniform size. Reches and Gazit used a for cell mobility and cell division. The growth of F-actin filaments

Fig. 7 (a) Scheme for the synthesis of actin-assembled Au nanowires; ref. 38. (b) Schematic diagrams of three methods for fabricating QD nanowires
using F-actin as a template (bottom) and fluorescence images of QD–actin nanowires (top); ref. 39.

This journal is © The Royal Society of Chemistry 2014 Chem. Commun., 2014, 50, 4148--4157 | 4155
 View Article Online

ChemComm Feature Article

can be controlled by thymosin and profilin. The former inhibits nanoparticles, allowing an electrical current to pass between
the polymerization process by binding to G-actin, while the the two surfaces for microelectronics applications.
latter promotes monomeric addition to the end of F-actin
filaments by binding to G-actin. As shown in Fig. 7a,38 after
the formation of F-actin filaments and the addition of gold
Conclusions
nanoparticles (Au NPs, 1.4 nm) and N-hydroxysuccinimide An overview of recent developments and applications for lever-
active ester, Au NP–F-actin composites could be obtained. The aging cell structures as building block nanomaterials (actin fila-
removal of ATP, Mg2+, and K+ degraded the composites into ments and MTs in this study) has been provided in this article.
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

Au NP–G-actin monomers, as confirmed by AFM and STEM Actin filaments interact to form braids, bundles, layers, and
(scanning transmission electron microscopy). Interestingly, columns whose architecture and mechanical properties regulate
these Au NP–G-actin monomers could be re-polymerized into and control cell shape. The capacity to prepare individual actin
Au NP–F-actin filaments. Subsequent enlargement of the Au filaments and subsequently use them for nanopatterning and
NPs by reduction of gold ions would eventually yield contin- functional nanowire manufacture has been illustrated here.
uous gold nanowires (Au NWs). Au NWs with lengths of 2–3 mm In regards to MTs, the proposed approach of in vitro kinesin-
and average heights of 80–100 nm, characteristics controlled by based transport using reconstructed suspended MTs may be a
metal deposition time, could be obtained via this self-assembly, promising method for the detection and characterization of
F-actin templating process. It is worth noting that, instead of molecular factors regulating intracellular transport and as a
direct coupling of Au NPs to G-actin monomers, coupling of Au vehicle for drug delivery. These results demonstrate that the
NPs to F-actin filaments and subsequent depolymerization of self-assembly process of actin filaments may have unanticipated
Au NP–F-actin filaments into Au NP–G-actin monomers are two industrial applications. In the near future, we look forward to
essential steps for successfully achieving Au NWs. This is continued improvements in these fields as ideas are shared and
because the direct coupling of Au NPs to G-actin monomers applications are made across disciplines.
blocked the linkage between two Au NP–G-actin monomers.
The self-assembly, F-actin templating process above could
be further expanded to prepare structures such as actin–Au
Acknowledgements
NW–actin and Au NW–actin–Au NW, and these patterned actin- We would like to thank the National Science Council of Taiwan
based Au NWs could be used as bio-nanotransporters. When for financially supporting this research under Contract No. NSC
these patterned actin-based Au NWs were placed onto a 101-2628-E-007-011-MY3, NSC 102-2221-E-007-031 (to C.-M.
myosin-coated surface, the nanowires were strongly linked with Cheng), and 101-2628-E-002-015-MY3 (to K. C.-W. Wu), along
myosin. After the addition of ATP–Mg2+ containing buffer, the with the National Health Research Institute (NHRI) of Taiwan
linkage between F-actin and myosin could be broken, thus (ME-102-PP-14; to K. C.-W. Wu), National Taiwan University
leading to the movement of the patterned actin-based Au NWs. (102R7842 and 102R7740; to K. C.-W. Wu) and the Center of
Yao et al. used F-actin as a template for the assembly of Strategic Materials Alliance for Research and Technology (SMART
water-soluble quantum dots (QDs) into nanowires.39 The unique Center), National Taiwan University (102R104100; to K. C.-W. Wu).
electronic and optical properties of QDs, together with advances
in QD synthesis and bio-functionalization paved the way
for potential QD nanowire applications in electro-optical and
Notes and references
biomolecular devices. The cross-linking approaches Yao et al. 1 (a) E. Katz and I. Willner, Angew. Chem., Int. Ed., 2004, 43, 6042;
proposed for the fabrication of biomolecule–nanoparticle hybrid (b) C. M. Niemeyer, Angew. Chem., Int. Ed., 2001, 40, 4128; (c) H. Gu,
K. Xu, C. Xu and B. Xu, Chem. Commun., 2006, 941; (d) R. Baron,
systems can be described as follows: (i) attach amino-QDs to the B. Willner and I. Willner, Chem. Commun., 2007, 323; (e) I. Willner,
carboxyl groups of F-actin via NHS–EDC cross-linker; (ii) use B. Basnar and B. Willner, FEBS J., 2007, 274, 302; ( f ) N. C. Seeman,
the bifunctional cross-linker sulfo-GMBS, which contains an Annu. Rev. Biochem., 2010, 79, 65.
2 (a) M. F. Krendel and E. M. Bonder, Cell Motil. Cytoskeleton, 1999,
amine-reactive sulfo-NHS ester at one end and a thiol-reactive 43, 296; (b) T. D. Pollard, L. Blanchoin and R. D. Mullins, Annu. Rev.
maleimide group at the other end; and (iii) QDs were conjugated Biophys. Biomol. Struct., 2000, 29, 545; (c) R. Heald and E. Nogales,
first with phallacidin by a EDC–NHS-mediated cross-linking J. Cell Sci., 2002, 115, 3; (d) C. I. Lacayo, Z. Pincus, M. M. VanDuijn,
C. A. Wilson, D. A. Fletcher, F. B. Gertler, A. Moqilner and
reaction and then incubated with F-actin for noncovalent bind- J. A. Theriot, PLoS Biol., 2007, 5, e233; (e) D. Nanba, F. Toki,
ing (Fig. 7b). In their study, they demonstrated that F-actin can N. Matsushita, S. Matsushita, S. Higashiyama and Y. Barrandon,
be used as a scaffold for the assembly of one-dimensional QD EMBO Mol. Med., 2013, 5, 640.
3 (a) K. C. Holmes, Nature, 2009, 457, 389; (b) T. Oda, M. Iwasa,
arrays or nanowires. Tightly packed on an actin filament, the T. Aihara, Y. Maéda and A. Narita, Nature, 2009, 457, 441.
QDs will be in close enough contact to transport electrons along 4 (a) P. Matsudaira, J. Bordas and M. H. Koch, Proc. Natl. Acad. Sci.
the filament and emit light. U. S. A., 1987, 84, 3151; (b) O. Pelletier, E. Pokidysheva, L. S. Hirst,
N. Bouxsein, Y. Li and C. R. Safinya, Phys. Rev. Lett., 2003,
Metal nanowire fabrication via actin-templating shows great 92, 148102.
potential as a new nanoscale machine that can be operated 5 (a) J. D. Kakisis, C. D. Liapis and B. E. Sumpio, Endothelium, 2004,
by biomolecules. In addition, actin columns in controlled 11, 17; (b) J. J. Wille, C. M. Ambrosi and F. C.-P. Yin, J. Biomech. Eng.,
2004, 126, 545; (c) R. Kaunas, P. Nguyen, S. Usami and S. Chien,
shapes and sizes can be self-assembled and joined to establish Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 15895.
connections, and the connections can be metallized with gold 6 H.-J. Hsu, C.-F. Lee and R. Kaunas, PLoS One, 2009, 4, e4853.

4156 | Chem. Commun., 2014, 50, 4148--4157 This journal is © The Royal Society of Chemistry 2014
 View Article Online

Feature Article ChemComm

7 J. Liao, L. Yang, J. Grashow and M. S. Sacks, Acta Biomater., 2005, 22 (a) N. Hirokawa, Science, 1998, 279, 519; (b) B. Alberts, A. Johnson,
1, 45. J. Lewis, M. Raff, K. Roberts and P. Walter, Mol. Biol. Cell, New York,
8 J. H.-C. Wang, P. Goldschmidt-Clermont, J. Wille and F. C.-P. Yin, 4th edn, 2002, p. 1616.
J. Biomech., 2001, 34, 1563. 23 K. Fujimoto, M. Kitamura, M. Yokokawa, I. Kanno, H. Kotera and
9 (a) M. Haqa, A. Chen, D. Gortler, A. Dardik and B. E. Sumpio, R. Yokokawa, ACS Nano, 2013, 7, 447.
Endothelium, 2003, 10, 149; (b) T. Lee and B. E. Sumpio, Biotechnol. 24 (a) T. Korten and S. Diez, Lab Chip, 2008, 8, 1441; (b) S. Taira,
Appl. Biochem., 2004, 39, 129; (c) K. Nishimura, W. Li, Y. Hoshino, Y.-Z. Du, Y. Hiratsuka, K. Konishi, T. Kubo, T. Q. P. Uyeda,
T. Kadohama, H. Asada, S. Ohgi and B. E. Sumpio, Am. J. Physiol.: N. Yumoto and M. Kodaka, Biotechnol. Bioeng., 2006, 95, 533;
Cell Physiol., 2005, 290, C812. (c) M. Bachand, A. M. Trent, B. C. Bunker and G. D. Bachand,
10 C.-M. Cheng and P. R. LeDuc, Adv. Mater., 2008, 20, 953. J. Nanosci. Nanotechnol., 2005, 5, 718; (d) M. C. Tarhan, Y. Orazov,
11 J. E. Smith, Vet. Pathol., 1987, 24, 471. R. Yokokawa, S. L. Karsten and H. Fujita, Lab Chip, 2013, 13, 3217.
Published on 22 January 2014. Downloaded by National Tsing Hua University on 20/10/2014 09:34:08.

12 J. Li, A. Shariff, M. Wiking, E. Lundberg, G. K. Rohde and 25 (a) C.-W. Wu, T. Aoki and M. Kuwabara, Nanotechnology, 2004,
R. F. Murphy, PLoS One, 2012, 7, e50292. 15, 1886; (b) C. W. Wu, T. Ohsuna, T. Edura and K. Kuroda, Angew.
13 J. Howard, A. Hudspeth and R. Vale, Nature, 1989, 342, 154. Chem., Int. Ed., 2007, 46, 5364.
14 (a) C. Z. Dinu, D. B. Chrisey, S. Diez and J. Howard, Anat. Rec., 2007, 26 (a) D. Popp, V. V. Lednev and W. Jahn, J. Mol. Biol., 1987, 197, 679;
290, 1203; (b) C. Z. Dinu, J. Opitz, W. Pompe, J. Howard, M. Mertig (b) C.-M. Cheng and P. R. LeDuc, J. Am. Chem. Soc., 2007, 129, 9546;
and S. Diez, Small, 2006, 2, 1090; (c) H. Hess, J. Clemmens, (c) C.-M. Cheng and P. R. LeDuc, Appl. Phys. Lett., 2008, 93, 174106;
C. Brunner, R. Doot, S. Luna, K. H. Ernst and V. Vogel, Nano Lett., (d) K. Morris and L. Serpell, Chem. Soc. Rev., 2010, 39, 3445; (e) M. Hazar,
2005, 5, 629; (d) H. Hess and V. Vogel, J. Biotechnol., 2001, 82, 67. R. L. Steward Jr., C.-J. Chang, C. J. Orndoff, Y. Zeng, M.-S. Ho,
15 (a) T. Kouyama and K. Mihashi, Eur. J. Biochem., 1980, 105, 279; P. R. LeDuc and C.-M. Cheng, Appl. Phys. Lett., 2011, 99, 233701.
(b) J. A. Cooper, S. B. Walker and T. D. Pollard, J. Muscle Res. Cell 27 M. Coles, W. Bicknell, A. A. Watson, D. P. Fairlie and D. J. Craik,
Motil., 1983, 4, 253; (c) N. Suzuki, H. Miyata, S. Ishiwata and Biochemistry, 1998, 37, 11064.
K. Kinosita Jr., Biophys. J., 1996, 70, 401; (d) D. A. Schafer, 28 Downloaded from the National Institute of Health; http://rsb.info.
P. B. Jennings and J. A. Cooper, J. Cell Biol., 1996, 135, 165. nih.gov/ij/download.html.
16 ATP-buffer is general actin buffer from Cytoskeleton; 29 L. Ionov, V. Bocharova and S. Diez, Soft Matter, 2009, 5, 67.
No.: BSA01. This buffer contains 5 mM Tris-HCl (pH 8.0) and 30 Y. K. Kryschenko, S. R. Seidel, A. M. Arif and P. J. Stang, J. Am. Chem.
0.2 mM CaCl2. Soc., 2003, 125, 5193.
17 This is a 10 solution that contains 500 mM KCl, 20 mM MgCl2 and 31 A. C. Reymann, J. L. Martiel, T. Cambier, L. Blanchoin, R. Boujemaa-
10 mM ATP. Paterski and M. Thery, Nat. Mater., 2010, 9, 827.
18 (a) M. L. Shelanski, F. Gaskin and C. R. Cantor, Proc. Natl. Acad. Sci. 32 (a) E. Braun, Y. Eichen, U. Sivan and G. Ben-Yoseph, Nature, 1998,
U. S. A., 1973, 70, 765; (b) R. C. Williams Jr. and J. C. Lee, Methods 391, 775; (b) S. Cui, Y. Liu, Z. Yang and X. Wei, Mater. Des., 2007,
Enzymol., 1982, 85, 376; (c) R. B. Vallee, Methods Enzymol., 1986, 28, 722; (c) S. H. Han and J. S. Lee, Langmuir, 2012, 28, 828.
134, 89; (d) J. Avila, H. Soares, M. L. Fanarraga and J. C. Zabala, Curr. 33 (a) F. Patolsky, Y. Weizmann, O. Lioubashevski and I. Willner,
Protoc. Cell Biol., 2008, 39, 3.29.1. Angew. Chem., Int. Ed., 2002, 41, 2323; (b) A. N. Sokolov, F. L. Yap,
19 (a) K. Rajagopal and J. P. Schneider, Curr. Opin. Struct. Biol., 2004, N. Liu, K. Kim, L. Ci, O. B. Johnson, H. Wang, M. Vosgueritchian,
14, 480; (b) K. Morris and L. Serpell, Chem. Soc. Rev., 2010, A. L. Koh, J. Chen, J. Park and Z. Bao, Nat. Commun., 2013, 4, 2402.
39, 3445. 34 J. Richter, M. Mertig, W. Pompe, I. Monch and H. K. Schackert, Appl.
20 (a) J. D. Cortese, B. Schwab, C. Frieden and E. L. Elson, Proc. Natl. Phys. Lett., 2001, 78, 536.
Acad. Sci. U. S. A., 1989, 86, 5773; (b) H. Miyata and H. Hotani, Proc. 35 M. Mertig, L. C. Ciacchi, R. Seidel, W. Pompe and A. De Vita, Nano
Natl. Acad. Sci. U. S. A., 1992, 89, 11547; (c) S. Li and A. F. Palmer, Lett., 2002, 2, 841.
Langmuir, 2004, 20, 4629; (d) S. Li and A. F. Palmer, Langmuir, 2004, 36 C. F. Monson and A. T. Woolley, Nano Lett., 2003, 3, 359.
20, 7917. 37 M. Reches and E. Gazit, Science, 2003, 300, 625.
21 H. Takatsuki, H. Tanaka, K. M. Rice, M. B. Kolli, S. K. Nalabotu, 38 F. Patolsky, Y. Weizmann and I. Willner, Nat. Mater., 2004, 3, 692.
K. Kohama, P. Famouri and E. R. Blough, Nanotechnology, 2011, 39 L. Yao, G. O. Andreev, Y. K. Reshetnyak and O. A. Andreev, Anal.
22, 245101. Bioanal. Chem., 2009, 395, 1563.

This journal is © The Royal Society of Chemistry 2014 Chem. Commun., 2014, 50, 4148--4157 | 4157