Shaoyu Chen1,2, Liangliang Yang3, Franco King-Chi Leung1, Takashi Kajitani4, Marc C A Stuart1, Takanori Fukushima4, Patrick van Rijn3, Ben L Feringa1,2. 1. Center for System Chemistry, Stratingh Institute for Chemistry, University of Groningen, AG Groningen 9747, The Netherlands. 2. Key Laboratory for Advanced Materials and Joint International Research Laboratory of Precision Chemistry and Molecular Engineering, Feringa Nobel Prize Scientist Joint Research Center, Frontiers Science Center for Materiobiology and Dynamic Chemistry, Institute of Fine Chemicals, School of Chemistry and Molecular Engineering, East China University of Science and Technology, Shanghai 200237, China. 3. Department of Biomedical Engineering, University Medical Center Groningen, University of Groningen, AV Groningen 9713, The Netherlands. 4. Laboratory for Chemistry and Life Science, Institute of Innovative Research, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan.
Abstract
Mimicking the native extracellular matrix (ECM) as a cell culture scaffold has long attracted scientists from the perspective of supramolecular chemistry for potential application in regenerative medicine. However, the development of the next-generation synthetic materials that mimic key aspects of ECM, with hierarchically oriented supramolecular structures, which are simultaneously highly dynamic and responsive to external stimuli, remains a major challenge. Herein, we present supramolecular assemblies formed by motor amphiphiles (MAs), which mimic the structural features of the hydrogel nature of the ECM and additionally show intrinsic dynamic behavior that allow amplifying molecular motions to macroscopic muscle-like actuating functions induced by light. The supramolecular assembly (named artificial muscle) provides an attractive approach for developing responsive ECM mimetic scaffolds for human bone marrow-derived mesenchymal stem cells (hBM-MSCs). Detailed investigations on the photoisomerization by nuclear magnetic resonance and UV-vis absorption spectroscopy, assembled structures by electron microscopy, the photoactuation process, structural order by X-ray diffraction, and cytotoxicity are presented. Artificial muscles of MAs provide fast photoactuation in water based on the hierarchically anisotropic supramolecular structures and show no cytotoxicity. Particularly important, artificial muscles of MAs with adhered hBM-MSCs still can be actuated by external light stimulation, showing their ability to convert light energy into mechanical signals in biocompatible systems. As a proof-of-concept demonstration, these results provide the potential for building photoactuating ECM mimetic scaffolds by artificial muscle-like supramolecular assemblies based on MAs and offer opportunities for signal transduction in future biohybrid systems of cells and MAs.
Mimicking the native extracellular matrix (ECM) as a cell culture scaffold has long attracted scientists from the perspective of supramolecular chemistry for potential application in regenerative medicine. However, the development of the next-generation synthetic materials that mimic key aspects of ECM, with hierarchically oriented supramolecular structures, which are simultaneously highly dynamic and responsive to external stimuli, remains a major challenge. Herein, we present supramolecular assemblies formed by motor amphiphiles (MAs), which mimic the structural features of the hydrogel nature of the ECM and additionally show intrinsic dynamic behavior that allow amplifying molecular motions to macroscopic muscle-like actuating functions induced by light. The supramolecular assembly (named artificial muscle) provides an attractive approach for developing responsive ECM mimetic scaffolds for human bone marrow-derived mesenchymal stem cells (hBM-MSCs). Detailed investigations on the photoisomerization by nuclear magnetic resonance and UV-vis absorption spectroscopy, assembled structures by electron microscopy, the photoactuation process, structural order by X-ray diffraction, and cytotoxicity are presented. Artificial muscles of MAs provide fast photoactuation in water based on the hierarchically anisotropic supramolecular structures and show no cytotoxicity. Particularly important, artificial muscles of MAs with adhered hBM-MSCs still can be actuated by external light stimulation, showing their ability to convert light energy into mechanical signals in biocompatible systems. As a proof-of-concept demonstration, these results provide the potential for building photoactuating ECM mimetic scaffolds by artificial muscle-like supramolecular assemblies based on MAs and offer opportunities for signal transduction in future biohybrid systems of cells and MAs.
Hierarchical supramolecular
polymeric structures are commonly found
in living systems, for example, the extracellular matrix (ECM), cell
membranes, and the cytoskeleton. They serve vital roles in key biological
functions, allowing for cell growth, division, and movement.[1−5] The ECM is a three-dimensional (3D) supramolecular network mainly
consisting of fibrous proteins and proteoglycans, present within all
tissues and organs, which provides not only important physical support
for the surrounding cells but also crucial biochemical and biomechanical
signals in a dynamic and spatiotemporal manner for tissue morphogenesis,
differentiation, and homeostasis.[6] Taking
inspiration from the essential roles of natural ECM in governing cell
functions in living systems, numerous synthetic polymers have been
developed for ECM mimics, ranging from 1D to 3D soft materials, providing
promising potential for tissue engineering and regenerative medicine,
such as stem cell-based therapies.[7−11] Due to the dynamic and adaptive nature of native ECM, supramolecular
polymers, including thermoplastic elastomers[7] on the basis of polyurethanes, bisurea, and ureidopyrimidinone motifs
as well as hydrogels, are attractive candidates, which have been applied
in vitro or in vivo for tissue regeneration.[7,9,12−15] Among the above biomaterials,
hydrogels are able to absorb up to 99% of water, allowing for the
encapsulation and diffusion of cells under physiological conditions,
which provide a 3D environment more closely resembling the native
ECM situation.[7−9] A special class of supramolecular hydrogels formed
by peptide amphiphile assemblies, reported by the Stupp group, is
among the most remarkable materials showing unique advantages of injectability,
biocompatibility, and biodegradability.[16−18] These peptide amphiphiles
are generally composed of alternating hydrophobic alkyl chains and
hydrophilic amino acid residues, which spontaneously form well-organized
assemblies ranging from nanoscale to macroscopic 3D isotropic or anisotropic
hydrogels.[16] Isotropic hydrogels, developed
from peptide amphiphiles combined with epitopes[19−22] or growth factors[23,24] or mimicked bioactive sequences,[25−29] have been used to mimic the ECM, showing possible
clinical applications in bone therapy;[27,30] brain,[31] kidney,[32] and central
nervous system injury;[33] cartilage regeneration;[34] and angiogenesis.[35,36] Compared to
isotropic hydrogels, anisotropic hydrogels with hierarchically high-oriented
structures are considered a particularly attractive class of ECM mimetic
scaffold for regenerative medicine.[37,38] They are important
for tissues with unidirectionally aligned structures, such as muscle
fibers, the spinal cord, bones, and parts of the brain.[37] Particularly, by gently shearing the annealed
solution of peptide amphiphiles into salt-containing media, highly
oriented noodle-like strings with an arbitrary length can be obtained
which perform as templates to induce the alignment of growth direction
of cells,[39] showing challenging applications
in regenerative medicine, for example, for blood vessels,[40] neural tissues,[41−43] bones,[44] cavernous nerves,[45] and muscle
tissues.[46] Except for the hierarchically
anisotropic supramolecular structure, the ideal ECM mimetic scaffold
for regenerative medicine also should be highly dynamic and responsive,
allowing for adaptiveness and spatiotemporal feedback control on cell
functions as a result of trigger signals or changes in the environment.[7,17] Only very few studies reported responsive ECM mimetic scaffolds
by using enzymatic cleavage or photocleavage to remove bioactive sequences,
typically being irreversible in nature.[47,48] In the approach
presented here, different advantageous material attributes are included,
however, to the best of our knowledge prior to this study, and no
single system has been found to be able to simultaneously satisfy
the following multiple requirements, that is, to provide biocompatible
supramolecular materials with hierarchically highly oriented structures,
high dynamics, as well as response to external stimuli in a non-invasive
manner and amplification of mechanical effects across the length scale
from molecular motions to macroscopic movements, which is likely a
significant step toward the development of next-generation dynamic
ECM mimetic scaffolds for tissue engineering and regenerative medicine.[7,17,18] It should be emphasized that
the synthetic systems mimic only certain aspects of ECM, that is,
in our design, biocompatibility with living cells and hierarchical
organization in water and while integrating this with our light-driven
motor function to allow non-invasive dynamic functions.Recently,
we have demonstrated the first example of a hierarchically
organized anisotropic supramolecular system comprising a motor amphiphile
(MA) in aqueous media, allowing for the development of
reversible photoactuating artificial muscles in water and in air.[49] These artificial muscles were prepared by self-assembly
involving the addition of MA aggregates (composed of
95% water) to a CaCl2 solution using a shear flow method
for alignment.[49−51] The unique hierarchically anisotropic supramolecular
structures enabled the amplification of molecular motions across length
scales. Note that our systems, in contrast to natural muscles, are
triggered by light as a non-invasive stimulus to induce macroscopic
photoactuation, allowing for light energy conversion to mechanical
motions. Because these MA artificial muscles meet multiple
requirements of potential next-generation ECM mimetic scaffolds, we
envisioned that they might be biocompatible and be applied in systems
controlling the fate of cells.Here, we present MAs, which are based on the second-generation
of the molecular motor core,[52−56] combined with a hydrophobic alkyl chain and hydrophilic chains with
various end groups, that is, analogues of charged groups in the native
ECM including carboxylate groups (MA), phosphite groups (MA), and sulfate groups (MA),
as shown in Figure . After investigating the effect of the functionalities on the molecular
isomerization process and supramolecular assembly in water, the MAs are used to prepare photoactuating artificial muscles
with highly oriented supramolecular structures. For the first time,
these structurally well-oriented artificial muscles are combined with
human bone marrow-derived mesenchymal stem cells (hBM-MSCs) to determine the cytotoxicity and mechanical motion in the presence
of cells. As a proof-of-concept demonstration, we explored the possibility
of photoactuating artificial muscles of MAs as ECM mimetic
scaffolds for mesenchymal stem cells. The prospects of transducing
mechanical signals to control cell functions in the next stage might
lead to potential applications such as in vitro pathology models for
studying complicated cell signaling response environments and to ultimately
provide stem cell-based therapies that address challenging medical
problems in the future, such as nerve tissue regeneration after spinal
cord injuries.
Figure 1
Schematic illustration of (a) reversible photoisomerization
and
thermal helix inversion of molecular MAs and (b) hierarchically
anisotropic supramolecular assembly structures of photoactuating artificial
muscles of MAs and the application of MA artificial muscles as ECM mimetic scaffolds for mesenchymal stem
cells.
Schematic illustration of (a) reversible photoisomerization
and
thermal helix inversion of molecular MAs and (b) hierarchically
anisotropic supramolecular assembly structures of photoactuating artificial
muscles of MAs and the application of MA artificial muscles as ECM mimetic scaffolds for mesenchymal stem
cells.
Results and Discussion
Molecular Design and Synthesis
Our earlier MAs,[49,50] for forming
artificial muscles, were designed
with a dodecyl chain attached to the upper half of a second-generation
molecular motor core and two carboxylate end groups connected with
two alkyl linkers to the lower half of the motor core. By considering
the common phosphorylated and sulfated groups in the native ECM, these
functional groups play important roles in the mineralization and providing
cation binding sites.[28,57] In this context, the end groups
of MA were extended from carboxylate groups (MA) to the analogues of naturally existing
charged groups (Figure ), that is, phosphite groups (MA) and sulfate groups (MA), which not only allow identifying the end-group effects on molecular
isomerization processes and assembled structures but also provide
biocompatible groups, allowing for potential applications in developing
photoactuating artificial muscles as photoresponsive ECM mimetic scaffolds
for mesenchymal stem cells. The synthetic pathways for the MAs are summarized in Scheme .
Scheme 1
Synthesis of MAs
Stator and rotor units, that is, compounds 1 and 5, were prepared by our reported procedures.[49,58] Compound 3 was obtained from thioxanthone 1 and alkyl bromide 2 by a Williamson ether formation
reaction in the presence of K2CO3 in DMF and
subsequently converted into thioketone 4 with Lawesson’s
reagent in toluene. Hydrazone 5 was oxidized in situ
with (diacetoxyiodo)benzene in DMF to provide the corresponding diazo-compound,
followed by the addition of freshly prepared thioketone 4, providing the corresponding episulfide 6. Desulfurization
with triphenylphosphine in toluene gave overcrowded alkene 7, which was hydrolyzed to provide diol 8. The new MAs were obtained via two different synthetic methods: functionalization
of 8 with diphenyl phosphite yielded MA,[59] while MA was obtained by a sulfate ester formation
of 8 with a sulfur trioxide pyridine complex and subsequent
hydrolysis.[60] The structures of all new
compounds were established by 1H, 13C nuclear
magnetic resonance (NMR) and high-resolution ESI mass spectrometry,
and detailed synthetic procedures and data are provided in the Supporting Information (SI: pages S3–S6,
Figures S9–S20).
Photoisomerization
and Thermal Helix Inversion of MAs
The photochemical and
thermal isomerization steps of MA and MA were examined
by 1H NMR and UV–vis spectroscopy.
Essentially identical proton signal shifts of Ha, Hb, and the methyl group adjacent to the stereogenic center
are observed in the 1H NMR spectra of CD2Cl2 solutions of MA and MA, and upon prolonged irradiation
with 365 nm light, photostationary states (PSS) with metastable/stable
isomer ratios of 85:15 are obtained for both MA and MA (Figures , S2 and Table S1). In the UV–vis
absorption spectra of CH3CN solutions of MA and MA, an increase in the absorption around 310 nm with a concomitant
decrease of the absorption band from 330 to 370 nm is observed upon
irradiating with 365 nm light, which is essentially identical to that
observed in MA,[49] indicating the isomerization from a stable configuration
to a metastable configuration (Figure ). Additionally, a clear isosbestic point at 327 nm
over the course of irradiation indicates that a selective photoisomerization
process occurs (Figure ). The transformation from the metastable isomer into the stable
isomer can be induced by heating. The thermal helix inversion processes
of MA and MA in CH3CN solutions were studied
by means of Eyring analysis (Figure S3 and Table S2). The activation parameters and half-life of MA and MA are presented in Table S2. For
example, the Gibbs free energy of activation (Δ‡G) of MA was
102.5 kJ mol–1, which corresponded to a half-life
(t1/2) of 27.9 h at 25 °C. In UV–vis
absorption spectra of MAs in water (Figure S4), similar spectra changes are observed, which are consistent with
the spectra performed in CH3CN (Figure ). The results of the photoisomerization
and thermal helix inversion of MA and MA are comparable
to MA,[49] indicating that selective molecular isomerization processes between
stable and metastable isomers are observed typical for the second
generation motors by light and subsequent heat stimuli. This provides
a solid basis for the further investigation of the amplification of
molecular motion to macroscopic photoactuation.
Figure 2
Selected regions of 1H NMR spectra (CD2Cl2, 25 °C, 500
MHz) of (a) MA and (b) MA in a stable
state (black) and a mixture (red) containing 85% metastable isomers
at PSS after irradiation (for the proton assignment, see Figure , for full spectra,
see Figure S2).
Figure 3
UV–vis
absorption spectra of (a) MA and (b) MA in CH3CN solutions (6.5 × 10–2 mM) before 365 nm
light irradiation (black), upon irradiation from
1 to 3 min (pink), and after irradiation to PSS (red).
Selected regions of 1H NMR spectra (CD2Cl2, 25 °C, 500
MHz) of (a) MA and (b) MA in a stable
state (black) and a mixture (red) containing 85% metastable isomers
at PSS after irradiation (for the proton assignment, see Figure , for full spectra,
see Figure S2).UV–vis
absorption spectra of (a) MA and (b) MA in CH3CN solutions (6.5 × 10–2 mM) before 365 nm
light irradiation (black), upon irradiation from
1 to 3 min (pink), and after irradiation to PSS (red).
Supramolecular Assembly and Photoactuation
Aqueous
solutions of MA or MA were prepared by direct dissolution into
double deionized water or tris-buffer (pH 7.4). The corresponding
aqueous solutions of MAs were heated at 80 °C for
30 min and subsequently cooled down to room temperature to afford
colorless transparent solutions. The critical aggregation concentration
(CAC) of MAs was measured by using a Nile Red fluorescence assay (NRFA),
which probes the internal hydrophobicity of the assemblies.[50,61,62] Significantly smaller blue shifts
of the emission wavelength of Nile Red are observed when aqueous solutions
of MAs are diluted below 0.01 mM, and the CACs of MA and MA are determined to be 0.76 and 1.51 μM, respectively
(Figure S5). Freshly prepared solutions
of MAs at a concentration of 3.9 mM, above the CAC, were
used to analyze the self-assembled structures by using cryogenic transmission
electron microscopy (cryo-TEM) to capture their solution-state morphologies. MA assembles into worm-like micelles
from hundreds of nanometers to micrometers in length and ∼5–6
nm in diameter (Figure a), which is reminiscent of the assemblies of MA.[49−51] Compared to the worm-like micelles
in MA and MA, significantly different micellar-type assemblies
(∼10–20 nm in diameter) are observed in MA (Figure b), indicating that the sulfate end groups (MA) results in a lower packing parameter
than the amphiphilic motors with the carboxylate (MA) and phosphite (MA) end groups (e.g., micelles: P ≤ 1/3,
worm-like micelles: 1/3 < P ≤ 1/2).[63−65] Considering the assemblies of MA,[49−51] the results imply a possible formation of photoactuating
artificial muscles by MA.
Figure 4
Cryo-TEM
images of aqueous solutions of (a) MA and (b) MA (concentration:
3.9 mM, above the CAC).
Cryo-TEM
images of aqueous solutions of (a) MA and (b) MA (concentration:
3.9 mM, above the CAC).The MA artificial
muscles were formed according to
our reported procedure.[49−51] Typically, an aqueous solution
of MAs (55 mM) was manually drawn into an aqueous solution
of CaCl2 (150 mM) from a pipette by a shear flow method,
allowing for the formation of a noodle-like artificial muscle with
an arbitrary length. As expected, noodle-like artificial muscles were
obtained from a MA solution,
while a direct dissolution of MA into the media was observed. Therefore, only MA was further investigated regarding photoactuation
and structural features, and both MA and MA were investigated
in cell culture studies. An artificial muscle of MA is prepared in a cuvette containing CaCl2 solution, which bends toward the light source from an initial
position of 0° to a flexion angle of 90°, upon 365 nm irradiation
for 60 s, with an actuating speed of 1.5 ± 0.2°/s (Figure ). Compared to the
actuating speed of the MA artificial
muscle (7.9 ± 0.4°/s) under identical conditions, different
actuation speeds can be obtained by changing the end groups, which
could be attributed to the effect of end groups on the molecular packing,
the calcium-ion binding and the degree of alignment of the assembly
structures.[49−51] Next, the structural properties of the MA artificial muscle, including the morphology
and structural orientation, were characterized by using scanning electron
microscopy (SEM), polarized optical microscopy (POM), and through-view
small-angle X-ray scattering (SAXS) technique, and the results are
shown in Figure .
Figure 5
Photoisomerization
step of MA and photoactuation
of an MA artificial muscle
after irradiation with 365 nm light for 60 s.
Figure 6
(a) SEM
and (b) POM images of a noodle-like artificial muscle of MA under crossed polarizers. The
POM images of MA artificial
muscle were tilted at 0, 45, 90, and 135° relative to the transmission
axis of the analyzer. A scale bar applied for all panels. For images
taken at other angles, see Figure S6. (c)
2D SAXS image of the MA artificial
muscle (inset: enlarged 2D image for q = 0.1–0.45
nm–1 at 25 °C). (d) 1D SAXS patterns of the MA artificial muscle of 2D SAXS
images in (c), showing the diffraction pattern in the direction perpendicular
to the long axis of the artificial muscle.
Photoisomerization
step of MA and photoactuation
of an MA artificial muscle
after irradiation with 365 nm light for 60 s.(a) SEM
and (b) POM images of a noodle-like artificial muscle of MA under crossed polarizers. The
POM images of MA artificial
muscle were tilted at 0, 45, 90, and 135° relative to the transmission
axis of the analyzer. A scale bar applied for all panels. For images
taken at other angles, see Figure S6. (c)
2D SAXS image of the MA artificial
muscle (inset: enlarged 2D image for q = 0.1–0.45
nm–1 at 25 °C). (d) 1D SAXS patterns of the MA artificial muscle of 2D SAXS
images in (c), showing the diffraction pattern in the direction perpendicular
to the long axis of the artificial muscle.The SEM image of an MA artificial
muscle in Figure a
shows arrays of unidirectionally aligned bundles of worm-like micelles,
and the POM images of a freshly prepared MA artificial muscle present uniform birefringence in
the direction of the long axis of the muscle (Figure b). The structural parameters and orientational
orders, that is, degree of alignment, of the MA worm-like micelles in the macroscopic noodle-like MA artificial muscle were obtained
by through-view SAXS measurements. In the 2D SAXS image of the MA artificial muscle, prepared on
a sapphire substrate at 25 °C (Figure c), a pair of spot-like scatterings is observed
in a smaller-angle region (q = 0.1–0.45 nm–1) (Figure c, inset), which is due to scattering from the unidirectionally
aligned bundles of worm-like micelles. The diffraction arcs with a d-spacing of 6.15 nm (Figure d) are attributed to the diffraction from the (001)
plane of a lamellar structure, which is constructed by the unidirectionally
aligned worm-like micelles of MA with ionic interaction between Ca2+ and phosphite of MA as interfibrillar interaction.
The layer spacing of the lamellar structure (c =
6.15 nm) of the MA artificial
muscle is longer than that observed in MA (c = 5.48 nm),[50] possibly due to a loose packing between MA worm-like micelles and Ca2+. The angular
dependency of the peak intensity of the diffraction from the (001)
plane, converted from the through-view 2D SAXS image of the MA artificial muscle, shows intensity
maxima at 0 and 180° (Figure c). The peak intensity of the diffraction from the
(001) plane was quantified by full-width half-maximum (fwhm) to obtain
an ∼100° azimuthal angle, in which a smaller azimuthal
angle represented a higher degree of unidirectional alignment. Given
that the MA artificial muscle
showed an ∼65° azimuthal angle,[50] the results indicated that a lower degree of unidirectional alignment
was present in the MA artificial
muscle. The results demonstrated that likely due to the effects of
end-groups in the molecular MAs on molecular packing and structural
orientation in supramolecular assembly, artificial muscles of MAs with different photoactuating speeds can be obtained.
Cell Culture and Post-photoactuation
To explore the
potential of MA artificial muscles as ECM mimetic scaffolds
for cell culture, hBM-MSCs were selected because of their
typical advantages toward differentiation possibilities and potential
future clinical translation.[33] Following
an identical preparation protocol for MA artificial muscles,
freshly prepared artificial muscles of MA and MA, as ECM mimetic
scaffolds, were placed in the 24-well plates containing growth medium
(0.5 mL), followed by evenly seeding hBM-MSCs into the
well plates with a density of 20,000 cells/well. hBM-MSCs were expected to randomly attach to the surface of MA artificial
muscles (for details procedure of the cell culture, see Supporting Information, SI page S8). After incubation
for 24 h, the cytotoxicity was analyzed by the live/dead assay by
using calcein-AM and ethidium homodimer-1 in PBS to stain hBM-MSCs. After staining, the MA artificial muscles were transferred
from the original 24-well plates to a glass plate or Petri dish, which
avoided significant background fluorescence from hBM-MSCs attached on the surface of the 24-well plates for the subsequent
cell observation under confocal laser scanning microscopy and fluorescence
microscopy. Live cells show green fluorescence due to the uptake of
calcein-AM via intracellular activity, while dead cells show red fluorescence
because of the entry of ethidium homodimer-1 through the damaged cell
membranes and its subsequent binding to nucleic acids. The predominantly
observed green fluorescence and in the absence of red fluorescence
strongly suggested an almost 100% cell viability of hBM-MSCs cultured in the artificial muscles of MA and MA (Figures and S7). Furthermore, it is shown that the live cells
are attached to the surface of the artificial muscle (Figures and S7). Using a direct contact method between hBM-MSCs and MA artificial muscles, MA artificial muscles,
as ECM mimetic scaffolds, showed no cytotoxicity, indicating a good
in vitro biocompatibility of MA artificial muscles.[66]
Figure 7
Cytotoxicity of artificial muscles of MA (top) and MA (bottom) for hBM-MSCs after 24 h of incubation,
determined
by a live/dead staining assay. The images were taken by confocal laser
scanning microscopy. Scale bar: 100 μm, applied for all panels.
The bar graphs (right) show the percentages of live cells in the fluorescent
images (for full images, see Figure S7).
Cytotoxicity of artificial muscles of MA (top) and MA (bottom) for hBM-MSCs after 24 h of incubation,
determined
by a live/dead staining assay. The images were taken by confocal laser
scanning microscopy. Scale bar: 100 μm, applied for all panels.
The bar graphs (right) show the percentages of live cells in the fluorescent
images (for full images, see Figure S7).To provide a deeper insight into the condition
of hBM-MSCs cultured on the surface of the MA artificial muscles,
the hBM-MSCs were seeded into 24-well plates (with a
density of 2500 cells/well) containing a growth medium and freshly
prepared MA artificial muscles, incubating for 1, 3,
and 5 d, and then 4′,6-diamidino-2-phenylindole (DAPI) and
tetramethylrhodamine isothiocyanate(TRITC)–phalloidin were
used to stain cell nuclei and F-actins of hBM-MSCs, respectively.
Subsequently, the corresponding MA artificial muscles
with adhered hBM-MSCs were transferred from the original
24-well plates to a glass plate and were observed using confocal laser
scanning microscopy (for details protocol, see Supporting Information, SI page S8). Generally, a small and
round cell shape is typically an indication of cells entering apoptosis,[67,68] whereas a well-spread cell shape of hBM-MSCs is commonly
quantified as being in a viable state.[69−71] As depicted in Figure a,b and Supporting
Information, Movie 1, cells spread with
well-defined actin stress fibers on the surfaces of MA artificial muscles, indicating that hBM-MSCs maintain
highly viable over prolonged incubation time. Due to overlapping fluorescence
between the strong background of MA artificial muscles
and the cell nuclei stained by DAPI, the change of cell F-actin area,
instead of the number of cell nuclei, was used to indicate the cell
proliferation. The cell F-actin area significantly increases upon
prolonged incubation time from 1 to 5 d, suggesting a possibility
of cell proliferation of hBM-MSCs on the surfaces of MA artificial muscles (Figure a,b). To further provide a quantitative analysis of
the change of F-actin area, the percentage of F-actin coverage was
calculated by dividing the area of F-actin by the total surface area
of MA artificial muscles (Figure c). The areas were measured by the software
of ImageJ.[69] Both in the MA and MA artificial muscles, the F-actin coverages show significant
increase from ∼15% (1 d culture) to ∼22% (3 d), and
to ∼40% (5 d), which not only suggests a highly viable state
of hBM-MSCs but also indicates a possibility of good
cell proliferation (Figure c).
Figure 8
F-actin of hBM-MSCs on the surface of artificial muscles
of (a) MA and (b) MA. The images were taken by confocal laser
scanning microscopy. Scale bar: 100 μm, applied for all panels. hBM-MSCs are stained using phalloidin for F-actin (red). The
bar graphs (c) show the changes of average percentages of cell F-actin
coverage on the surface of artificial muscles of MA (top) and MA (bottom) over various culture days, indicating a possibility
of cell proliferation. The percentage of cell F-actin coverage was
detected by dividing the area of F-actin by the total surface area
of MA artificial muscles. A value of 100% indicates that
the artificial muscle is completely covered. Data are shown as mean
(±) standard deviation (SD), and **P < 0.01,
***P < 0.001, ****P < 0.0001.
F-actin of hBM-MSCs on the surface of artificial muscles
of (a) MA and (b) MA. The images were taken by confocal laser
scanning microscopy. Scale bar: 100 μm, applied for all panels. hBM-MSCs are stained using phalloidin for F-actin (red). The
bar graphs (c) show the changes of average percentages of cell F-actin
coverage on the surface of artificial muscles of MA (top) and MA (bottom) over various culture days, indicating a possibility
of cell proliferation. The percentage of cell F-actin coverage was
detected by dividing the area of F-actin by the total surface area
of MA artificial muscles. A value of 100% indicates that
the artificial muscle is completely covered. Data are shown as mean
(±) standard deviation (SD), and **P < 0.01,
***P < 0.001, ****P < 0.0001.In addition to the excellent viability of hBM-MSCs cultured in the MA artificial muscles,
it should be
emphasized that MA artificial muscles with adhered hBM-MSCs maintain their photoactuation function (Movies 2, 3, Figures and S8). During the macroscopic photoactuation of MA artificial muscles with adhered hBM-MSCs on
the surface, the in situ photoactuating process was monitored under
fluorescence microscopy and simultaneously by digital movies (Figure and Movie 2, Figure S8 and Movie 3). The method allows showing
the photoactuating movement accompanied with a cell visualization
(live cells stained by using calcein-AM). The MA artificial muscle with adhered hBM-MSCs bends toward the light source from an initial position of 0°
to a flexion angle of 21° upon 365 nm photoirradiation for 30
s (Figure ). Upon
prolonged photoirradiation to 60 s, an increased flexion angle to
40° is obtained (Movie 2). Compared
to MA, due to a loose packing
between worm-like micelles of MA and Ca2+ as well as a lower degree of unidirectional
alignment (Figure ), a slower photoactuating speed of MA artificial muscles with adhered hBM-MSCs is observed,
that is, the MA artificial
muscle bends from an initial position of 0° to a flexion angle
of 21° upon 365 nm photoirradiation for 200 s (Movie 3, Figure S8), again indicating
that the photoactuating speeds of MA artificial muscles
with adhered hBM-MSCs can be tuned by the end-groups.
The results demonstrated that MA artificial muscles,
as responsive ECM mimetic scaffolds for hBM-MSCs, showed
no cytotoxicity. Particularly important, MA artificial
muscles with adhered hBM-MSCs maintain their photoactuation
function with tunable actuation speed, demonstrating the capability
of photoenergy conversion into mechanical actuation in the presence
of hBM-MSCs. We demonstrated that a multifunctional synthetic–biological
system in water, comprising artificial muscles of MAs and mesenchymal stem cells, is capable of converting light energy
into mechanical actuation from the molecular level to macroscopic
dimensions. As a proof-of-concept study, these results provide attractive
applications of photoactuating artificial muscles of MAs as ECM mimetic scaffolds for hBM-MSCs, suggesting opportunities
for the transduction of a mechanical actuation signal to control cell
functions in the future.
Figure 9
Snapshots showing (a) macroscopic photoactuation
(scale bar: 5.0
mm) of an MA artificial muscle
with adhered hBM-MSCs on the surface and (b) in situ
observation of photoactuating movement accompanied with cell visualization
under fluorescence microscopy upon 365 nm light irradiation for 30
s (scale bar: 100 μm), demonstrating the maintaining of photoactuating
function of MA artificial muscles after their application
as ECM mimetic scaffolds for mesenchymal stem cells.
Snapshots showing (a) macroscopic photoactuation
(scale bar: 5.0
mm) of an MA artificial muscle
with adhered hBM-MSCs on the surface and (b) in situ
observation of photoactuating movement accompanied with cell visualization
under fluorescence microscopy upon 365 nm light irradiation for 30
s (scale bar: 100 μm), demonstrating the maintaining of photoactuating
function of MA artificial muscles after their application
as ECM mimetic scaffolds for mesenchymal stem cells.
Conclusions
Molecular MAs with variously
charged end groups, that
is, carboxylate groups (MA),
phosphite groups (MA), and
sulfate groups (MA), were designed
to form artificial muscles. Both artificial muscles of MA and MA allow fast photoactuation due to the amplification of molecular
motor motion along length scales. This is based on the distinct isomerization
processes of the motor units and the hierarchically anisotropic assembly
structures in water, as confirmed by UV–vis absorption and
NMR spectroscopy, electron microscopy, and X-ray diffraction. Taking
artificial muscles of MA and MA as ECM mimetic scaffolds, they
show no cytotoxicity, and particularly important, MA artificial
muscles with adhered hBM-MSCs still can be actuated by
photoirradiation. This study shows the feasibility of developing the
next generation of ECM mimetic scaffolds by using artificial muscles
of MAs with highly oriented supramolecular structures
and the ability to convert light energy into mechanical actuation
in biological systems. Following this proof of concept, the possibility
of using MA artificial muscles as responsive scaffolds
for mesenchymal stem cells, using visible light-driven artificial
muscles is envisioned. This will enable to systematically investigate
cell properties and explore the transduction of mechanical actuation
signal to control the differentiation of mesenchymal stem cells, which
is a part of our ongoing program with future prospects toward responsive
materials for regenerative medicine.
Authors: Rashad Mammadov; Busra Mammadov; Sila Toksoz; Bahri Aydin; Ramazan Yagci; Ayse B Tekinay; Mustafa O Guler Journal: Biomacromolecules Date: 2011-09-07 Impact factor: 6.988
Authors: Jiawen Chen; Franco King-Chi Leung; Marc C A Stuart; Takashi Kajitani; Takanori Fukushima; Erik van der Giessen; Ben L Feringa Journal: Nat Chem Date: 2017-12-04 Impact factor: 24.427
Authors: Wei Ji; Zaida Álvarez; Alexandra N Edelbrock; Kohei Sato; Samuel I Stupp Journal: ACS Appl Mater Interfaces Date: 2018-11-26 Impact factor: 9.229
Authors: Ramille N Shah; Nirav A Shah; Marc M Del Rosario Lim; Caleb Hsieh; Gordon Nuber; Samuel I Stupp Journal: Proc Natl Acad Sci U S A Date: 2010-02-01 Impact factor: 11.205
Authors: Stephan Sinn; Liulin Yang; Frank Biedermann; Di Wang; Christian Kübel; Jeroen J L M Cornelissen; Luisa De Cola Journal: J Am Chem Soc Date: 2018-02-06 Impact factor: 15.419
Authors: Franco King-Chi Leung; Tobias van den Enk; Takashi Kajitani; Jiawen Chen; Marc C A Stuart; Jeroen Kuipers; Takanori Fukushima; Ben L Feringa Journal: J Am Chem Soc Date: 2018-12-04 Impact factor: 15.419
Authors: Faifan Tantakitti; Job Boekhoven; Xin Wang; Roman V Kazantsev; Tao Yu; Jiahe Li; Ellen Zhuang; Roya Zandi; Julia H Ortony; Christina J Newcomb; Liam C Palmer; Gajendra S Shekhawat; Monica Olvera de la Cruz; George C Schatz; Samuel I Stupp Journal: Nat Mater Date: 2016-01-18 Impact factor: 43.841