Facile fabrication of high-definition hierarchical wrinkle structures for

Subscriber access provided by Stockholm University Library

Biological and Medical Applications of Materials and Interfaces

Facile fabrication of high-definition hierarchical wrinkle structures for investigating the geometrysensitive fate commitment of human neural stem cells Jieung Baek, Woo-Bin Jung, Younghak Cho, Eunjung Lee, GeunTae Yun, Soo-Yeon Cho, Hee-Tae Jung, and Sung Gap Im ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.9b03479 • Publication Date (Web): 22 Apr 2019 Downloaded from http://pubs.acs.org on April 22, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Facile Fabrication of High-Definition Hierarchical Wrinkle Structures for Investigating the Geometry-Sensitive Fate Commitment of Human Neural Stem Cells Jieung Baek†,〦, Woo-Bin Jung†,‡,〦, Younghak Cho†, Eunjung Lee†, Geun-Tae Yun†,‡, Soo-Yeon Cho†,‡, Hee-Tae Jung†,‡, * and Sung Gap Im†,‡,* † Department

of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology,

291 Daehak-ro, Daejeon 34141, Korea ‡

KAIST Institute for Nanocentury, 291 Daehak-ro, Daejeon 34141, Korea

〦These

authors contributed equally to this work

*e-mail:

[email protected], [email protected]

Keywords: hierarchical wrinkle, multiscale wrinkle, neural stem cell (NSC), human neural stem cell (hNSC), mechanotransduction, differentiation, neurogenesis

1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT Since neural stem cells (NSCs) interact with biophysical cues from their niche during development, it is important to understand the biomolecular mechanism how the NSCs process these biophysical cues to regulate their behaviors. In particular, anisotropic geometric cues in micro-/nanoscale have been utilized to investigate the biophysical effect of the structure on NSCs behaviors. Here, a series of new nanoscale anisotropic wrinkle structures with the wide range of wavelength scales (from 50 nm to 37 μm) was developed to demonstrate the effect of the anisotropic nanostructure on the fate commitment of NSCs. Intriguingly, two distinct characteristic length scales promoted the neurogenesis. Each wavelength scale showed a striking variation in terms of dependency on the directionality of structures, suggesting the existence of at least two different ways in processing of anisotropic geometries for neurogenesis. Furthermore, combined effect of the two distinctive length scales was observed by employing hierarchical multiscale wrinkle structures with two characteristic neurogenesispromoting wavelengths. Taken together, the wrinkle structure system developed in this study can serve as an effective platform to advance the understanding of how cells sense anisotropic geometries for their specific cellular behaviors. Furthermore, this could provide clues for improving nerve regeneration system of stem cell therapies.

2


Page 2 of 30

Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Recently, it has been found that anisotropic contact for cell alignment plays a critical role in controlling the various cellular behaviors including cytoskeleton reorganization, nucleus gene expression, and extracellular matrix (ECM) remodeling1-3. In particular, neural stem cells (NSCs)4, critical for learning and memory in mammals, can alter their fate choice in response to the anisotropic geometrical input for the cell alignment, such as topographies for line contact5-8. Human neural stem cells (hNSCs) exhibited promoted neurogenesis on nanoscale anisotropic line patterns via facilitated focal adhesion (FA) formation7, 9. For example, Yang et al. suggested a model for the enhanced differentiation of hNSCs on nano-grating structures by the activation of integrin-mediated mechanotransduction and intracellular signaling pathways, mitogen-activated protein kinase (MAPK)-extracellular signal-regulated kinase (ERK)7. On the other hand, in the case of the grating structures with extremely low contact width ( 4 cm2). In contrast, 1D wrinkle with 1.4 μm wavelength is oriented in a single direction, resulting in an aligned FFT result. In order to clarify the amplitude and the aspect ratio of the fabricated wrinkle structures, focused ion beam (FIB) SEM was 6


Page 6 of 30

Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conducted. Cross-sectional SEM images of each sample show layered wrinkle structure. Both nanoscale 100 nm and microscale 1.4 μm wrinkle structures showed a low aspect ratio ~ 1/2, with the estimated amplitudes ~50 nm and ~820 nm for 100 nm and 1.4 μm wrinkles, respectively. In addition to 100 nm and 1.4 μm wrinkles, a wide range of wavelength was achieved by controlling the skin layer thickness (Figure 1d). In general, the wrinkle wavelength (𝜆) is proportional to the skin layer thickness, λ ≈ 2πh(𝐸𝑠/3𝐸𝑏)1/3 (𝐸 = 𝐸 (1 ― 𝜈2)), where Es, Eb and υ are the moduli of the skin layer and bulk substrate, and the Poisson’s ratio, respectively. Wavelength of nanoscale wrinkle was tunable by controlling the power and incident angle of Ar plasma. As Ar plasma power increases from 0.1 kV to 1 kV, the penetration depth of Ar plasma into PS film increases and wavelength increases from ~ 100 nm to ~ 450 nm. Between 2D and 1D wrinkles, there is no significant difference in the resultant wrinkle size. Wavelength under 100 nm was obtained by changing incident angle of Ar plasma, where the change in the incident angle of Ar plasma (100 V) from 90 ° to 30 ° resulted in the variation of wavelength from ~ 100 nm to ~50 nm (Figure S1 and Table S1 in Supplementary Information). Changing the plasma power and incident angle resulted in corresponding skin layer thickness because the penetration depth of Ar plasma is proportional to power and angle21. Different from nanoscale wrinkles, the wavelength of microscale wrinkle could be tuned by controlling the PVP concentration (CPVP). As CPVP increases from 1 wt% to 20 wt%, the wavelength increases from ~ 1.4 μm to ~ 37 μm (Figure S1 in Supplementary Information).

7



Figure 1┃Fabrication of nano- and microscale wrinkles with controlled orientation. Schematic illustration of a, nano- and b, microscale wrinkles. c, Top and cross-sectional SEM images of 2D and 1D wrinkle in each feature dimension. d, Wavelength of 2D and 1D wrinkles with different conditions depending on two different methods of skin layer preparation: Ar ion plasma treatment (left) and PVP coating (right).

Neurogenesis dependent on the length scale and directionality of geometrical contact Having established the wrinkle structures with the controlled wavelengths from nano- to micro-scales, we next determined whether the fate commitment of hNSCs is altered in response to the length scale variation and the directionality of the structures. For this purpose, the fate commitment of hNSCs cultured on each wrinkle structure was monitored by co-staining for Tuj1 and GFAP, and compared to that on flat PS film (Figure 2a, b, and Figure S2 in Supplementary Information). Interestingly, a clear difference was observed in the tendency of number percentage of Tuj1-positive cells with respect to the directionality of the wrinkle pattern while maintaining the same wrinkle wavelength. Compared to that on flat PS surface control, distinctively enhanced neurogenesis of hNSCs was observed on the 1D structures at two wavelength regions: at 100 nm and the wavelengths over 1.4 μm (Figure 2c). On the other 8


Page 8 of 30

Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


hand, 2D structures exhibited increased level of neurogenesis only at 100 nm, and no statistically meaningful increase in neurogenesis was detected in the wavelength region of over 1.4 μm (Figure 2d). Taken together, these results apparently indicate that the promoted neurogenesis on the wrinkle structures with the wavelength of 100 nm does not depend on the directionality of contact, whereas that on the structures over 1.4 μm wavelength is directionality-dependent. In addition, a statistically meaningful difference was detected in the relative proportion of other cells between 1D structure with 1.4-μm wavelength (1D-1.4 μm, hereafter) and 2D structure with 1.4-μm wavelength (likewise, 2D-1.4 μm) (Figure 2e). The group designated as ‘other cells’ represents the cells positive for only DAPI or for both Tuj1 and GFAP. Considering the previous reports referring that the cells co-labeled for both Tuj1 and GFAP are neuroprogenitors22 and that the proportion of the fate commitment to oligodendrocyte is negligible compared with neurogenesis or astrogenesis23, most of the cells in ‘other cell’ groups would be composed of undifferentiated cells. The relative proportion of other group cells on 1D-1.4 μm structure was lower than that on 2D-1.4 μm structure with a p value of 0.001, which shows that the contact-guided hNSC alignment influenced the total differentiation of hNSCs on the 1.4 μm wrinkle structure. Considering that the total differentiation involves both neurogenesis as well as astrogenesis and differentiation into oligodendrocytes, this result corresponds well with the previous report illustrating that the enhanced neurogenesis resulted from the promoted total differentiation of hNSCs on linear structures through the activation of MAPK-ERK pathway7, rather than from a change in selective mannor between neurogenesis and astrogenesis. On the other hand, no palpable difference was detected in the other group cell fractions between 1D structure with 100-nm wavelength (1D-100 nm) and 2D structure with 100-nm wavelength (2D-100 nm), which also did not show any apparent cell alignment along 9



the wrinkle orientation.

Figure 2┃Neurogenesis of hNSCs is dependent on the scale and the orientation of wrinkle structures. Confocal images showing the hNSCs differentiated for 4 days on a, flat PS surface, and b, the wrinkle structures (1D-100 nm, 1D-1.4 μm, 2D-100 nm, and 2D-1.4 μm, respectively). The cells were stained for Tuj1 (green), GFAP (red), and DAPI (blue), respectively. Scale bars: 100 μm. Number percentage of Tuj1-positive cells on c, 1D and d, 2D wrinkle structures with each wavelength scale from 50 nm to 37 μm. Red dot lines represent mean value of Tuj-positive cells (%) on flat PS control. p values were calculated through one-way ANOVA, followed by Tukey post hoc analysis. (n = 4, *p < 0.05 vs. flat PS control group). e, The relative proportion of each marker-positive (Tuj1 and GFAP) cells on each substrate (flat PS surface, and the wrinkle structures with the wavelength of 1D-100 nm, 2D-100 nm, 1D-1.4 μm, and 2D-1.4 μm, respectively). All the data represent mean ± standard deviation.

Suppressed cellular adhesion of hNSCs on the wrinkle structures with 100-nm wavelength In order to elucidate the reason for the existence of specific wavelength regions promoting the neurogenesis, cellular morphology of hNSCs on the wrinkle structures were monitored using SEM (Figure 3a). The cells on flat PS surface showed directionless, stretched cellular morphology, whereas the cell spreading was substantially suppressed on 1D-100 nm and 2D100 nm wrinkle structures. No appreciable alignment of cells along the wrinkle direction was observed, presumably due to the far smaller size of the wavelength than that of the whole cell body10. On the other hand, the size of 1.4-μm wrinkle structure was comparable to that of cell 10


Page 10 of 30

Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


body, and morphologies of the cells on both 1D-1.4 μm and 2D-1.4 μm structures showed higher morphological dependency on the 1.4-μm wrinkle structures than 100-nm wavelength structures. The cells on 1D-1.4 μm structure exhibited more elongated morphologies than those on 2D-1.4 μm structures. To understand how the wavelength scales influence the cellular responses, we performed focal adhesion (FA) assay, and checked the FA area density of single cells on each substrate (Figure 3b). Interestingly, a substantial decrease in the FA area density was observed on both 1D-100 nm and 2D-100 nm wrinkle structures with the decrease more than 60% from that on flat PS surface. Along with this observation, a slight reduction in vinculin expression was also monitored on the 100-nm wrinkle structure by western blot analysis (Figure 3c). Considering that the average threshold distance between each integrin-binding site for clustering of cells is around 70 nm24-25 and that the cells usually contact on wrinkle top (Figure S3 in Supplementary Information), both 1D and 2D 100-nm wrinkle structures, slightly larger than the average clustering site distance, would restrict the clustering of cells. The reduced vinculinpositive area density (Figure 3b) on both 100-nm structures also supports the hypothesis of the restricted FA clustering among the adjacent adhesion sites on the 100-nm structures. Immunostaining image analysis also supports the observations above, where the actin fiber formation was reduced significantly with limited adhesion clustering on both 1D-100 nm and 2D-100 nm structures compared to that on flat PS surface (Figure 3d). These results correlate well with a previous report showing that nanostructures can restrict cellular adhesion of NSCs and neural progenitor cells (NPCs)26. The nano-wrinkle structure induces the restricted clustering and formation of FA complexes, resulting in the decrease in the formation of stress fibers, which in turn, causes the traction force decrease, and triggers the change of cellular program associated with the neuronal differentiation26. Such a tendency is consistent with 11



previous observation5, where the extremely limited cellular contact showed enhanced neurogenesis through the limited activation of Rho GTPase and crosstalk between Rho A27-28 and YAP29-30.

Promoted Cytoskeletal alignment of hNSCs by wrinkle structures with 1.4-μm wavelength Another important feature in topographical sensing of hNSCs is deeply related with the alignment of cells along the structure directionality (Figure 3e, and Figure S4 in Supplementary Information). The guided alignment was determined by checking the % neurites with the angles lower than 10° in the wavelength region. A gradual increase in the cellular alignment was observed from 1D-50 nm to 1D-1.4 μm wrinkle structures, where the enhancement of neurogenesis also had been observed. In contrast, the cells on 2D structure exhibited randomly oriented neurites just as they are on flat surface (Figure S4 in Supplementary Information). These results correlate well with the aforementioned results, where the promoted neurogenesis was dominantly observed only on 1D structure but not in 2D structure (Figure 2c, d), showing a close relationship between the neurite orientation and the neurogenesis of hNSCs. Furthermore, the difference in neurite orientation between 1D and 2D structures resulted in a distinct difference in actin morphologies (Figure 3f). Highly oriented cellular morphology along the 1D-1.4 μm wrinkle structure enabled the formation of more aligned, clearer apical stress fibers along the wrinkle structure. On the other hand, the cells on 2D structure showed no appreciable organization of F-actin structures in apical planes. Such tendencies were quantified as the actin anisotropy ratio, where more polarized actin yields higher anisotropy ratios, showing a markedly higher value on 1D-1.4 μm structure than those 12


Page 12 of 30

Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


on flat PS surface and 2D-1.4 μm wrinkle structure (Figure 3g). This result indicates that the alignment along the anisotropic structure promotes actin formation. In addition, a slight increase in the activation of FAK (phosphorylated-FAK) was observed with the wavelength increase from 100 nm to 37 μm in 1D wrinkle structures (Figure 3h). This result is also consistent with the previous reports that showed the enhanced differentiation of hNSCs on nano-line structures with the activation of integrin-mediated mechanotransduction and intracellular signaling pathways related with FAK7. Taken together, the observations above strongly infers the existence of at least two different processing ways that cells sense the anisotropic geometries in size-dependent manner for their fate commitment of promoted neurogenesis. From the wrinkle structure with 1.4-μm wavelength, which is comparable to the size of axon diameter, a marked change in morphological alignment along to the contact was observed, followed by facilitated actin bundle formation, which clearly confirms its dependency on the orientation of the wrinkle structures with the size higher than 1.4 μm. In contrast, 100-nm wrinkle structures, which did not show any distinguishable dependency on the orientation of the anisotropic structures, only showing the restricted FA clustering formation (Figure 3i and Figure S5 in Supplementary Information).

13



Figure 3 ┃ Both size and orientation of wrinkle structures modulate the cellular mechanosensing process. a, SEM images of hNSCs on each substrate. Scale bar = 4 μm. b, Area density of FA calculated with the cells on the wrinkle structures with different wavelengths (50, 100, 300, 700, 1000, 1400, and 20000 nm for 1D and 2D structures, respectively) (n > 10 cells per each group, *p < 0.05 vs. flat surface group). c, Western blot for vinculin in hNSCs on each wrinkle structure. d, Representative fluorescence images showing hNSCs adhered on flat PS surface and wrinkle structures (1D-100 nm and 2D-100 nm) stained for actin (red), vinculin (green), and nucleus (blue); scale bar = 20 μm. e, Number percentage of neurites (θ 170 neurites per each group, *p < 0.05 vs. flat surface group). f, Representative fluorescence images for the cells on 1D-1.4 μm and 2D-1.4 μm wrinkles showing actin (red), vinculin (green), and nucleus (blue); scale bar = 10 μm. g, Quantified anisotropy ratios. More polarized actin yields higher anisotropy ratios (n > 9 cells per group). h, Western blotting conducted to compare the expression of phosphorylated FAK [pFAK(Y397)] in hNSCs cultured on flat PS surface, and the wrinkle structures with the wavelength of 100 nm, 1.4 μm, and 37 μm. i, A schematic illustration of suggested FA clustering and cytoskeleton formation in hNSCs on each substrate (flat PS surface, 100-nm and 1.4-μm wrinkle structures). All the data represents the mean ± s.e.m. p values were calculated through one-way ANOVA, followed by Tukey post hoc analysis. *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001. 14


Page 14 of 30

Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Enhanced neurogenesis on multiscale wrinkle including 100-nm and 1.4-μm wavelengths Based on our results demonstrating the two distinctive length scales for promoting neurogenesis on anisotropic geometry, we next checked the effect of these two length scales especially in case when the two length scales are presented together for cellular contact. For this purpose, we constructed a multiscale wrinkle structure containing both the two anisotropic structures with the wavelengths of 100 nm and 1.4 μm. Multiscale wrinkle was fabricated by hierarchical wrinkling method using a sacrificial skin layer (Figure 4a). After the generation of nanoscale wrinkle, PVP was coated onto the first-generated wrinkle as a sacrificial skin layer. By heating above 130 ℃ (> Tg = 120 ℃ of PS film) followed by the PVP layer removal with ethanol, hierarchical wrinkle composed of nano- and microscale wrinkles was formed successfully. Our method of creating hierarchical wrinkles enabled the individual adjustment of the size of nano- and micro-scale wrinkles. In addition, not only the size but also the directionality can also be controlled independently to fabricate various combinations of hierarchical wrinkle directions. Figure 4b and Figure S6 in Supplementary Information show the top and cross-sectional SEM images of hierarchical wrinkle with 100 nm and 1 μm wrinkles, where yellow line is the secondly generated (G2) wrinkle. Note that the G2 wavelength of multiscale wrinkles is changed slightly from that of single 1.4 μm wrinkle, which is likely caused by different thickness of PVP layer because of the underneath G1 wrinkles during spin coating. The G2 wavelength of multiscale wrinkle structure can be adjusted precisely by controlling the PVP concentration, enabling to tune the G2 wrinkle wavelength to 1.4 μm (Figure 4c and Figure S7 in Supplementary Information). The SEM images for the hNSCs adhered on each structure exhibited that the cells on the multiscale structure sustained its aligned morphologies guided by the contact line derived from 15



1.4-μm wavelength structure (Figure 4d). However, the adhesion morphology characteristics was clearly different from that on the structure with only single 1.4-μm wavelength. The cell on multiscale structure showed a limited cell spreading along the crest site of 1.4-μm wave in the hierarchical structure, which is seemingly due to the inhibitory role of the cellular contact with the 100-nm wrinkle structure on the cellular adhesion formation that we have observed previously (Figure 3a). Thus, the hNSCs on the hierarchical structure showed an aligned cellular morphology with limited cellular adhesion, affected by both two wavelengths. Intriguingly, the cells on the hierarchical structure showed a far enhanced neurogenesis, compared with all the other anisotropic structures that have shown higher neurogenesis up to approximately 80 % than flat PS surface: 1D-100 nm, 2D-100 nm, and 1D-1.4 μm (Figure 4e, f). The combined enhancement of the neurogenesis in the multiscale wrinkles led us to investigate the reason why the two different wavelength scales (100 nm and 1.4 μm) enable the more enhanced neurogenesis than when they were presented separately. In order to figure this out, FA formation and cytoskeleton were observed. Although the cells on hierarchical structure exhibited aligned morphologies as compared with the cells on flat surface and 1D-100 nm structure did, the hierarchical structure did not facilitate cytoskeleton alignment as the 1D-1.4 μm structure did, with lower number percentage of aligned neurites (θ 3 cells per group, *p < 0.05 vs. 1D-1.4 μm group). i, Area density of focal adhesion calculated with the cells on the wrinkle structures. j, A schematic illustration of focal adhesion clustering and cytoskeleton formation in hNSCs on hierarchical structure.

18


Page 18 of 30

Page 19 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


CONCLUSION In conclusion, anisotropic wrinkle structures with the vast wavelength range (from 50 nm to 37 μm) were utilized to investigate the geometry-dependent hNSC differentiation. The wrinkle system is basically very easy and cheap to fabricate surface pattern in the large area. Especially, the wrinkle system developed in this study, which can be made in a wide range of size control and capable of creating a hierarchical structure, is a very useful system for biology research requiring various scale adjustment and combination. Interestingly, neurogenesis was promoted at two distinctive wavelength regions: 100 nm and the region above 1.4 μm. The directionality of the wrinkle structure was also a critically important factor to influence the hNSC neurogenesis, which was dependent upon the wrinkle size. Structures with the 100-nm wavelength were turned out to restrict FA clustering and cytoskeleton formation, regardless of the structure orientation. On the other hand, the cells on the 1.4-μm structure showed a prominent cytoskeletal orientation with enhanced expression of pFAK. Particularly, a clear dependency on directionality was observed in this wavelength scale, where the cytoskeletal alignment was detected only on 1D-1.4 μm but not on 2D-1.4 μm structure. Taken together, our findings suggest that there would be at least two different geometry sensing process for promoting neurogenesis of hNSCs. Furthermore, multiscale structure containing both the two wavelengths (1D-100 nm and 1D-1.4 μm) was constructed and employed to investigate the combined effect of both two wrinkle sizes. The cells showed a further enhanced neurogenesis on the multiscale structures as compared with each single wrinkle structure, likely due to the even further restriction of FA clustering than the 100-nm single structure. Our study with the vast scale range of wrinkle structure platform would advance the understanding the biomolecular mechanism how the cells process the geometrical inputs and could give a clue for improving the regulation of nerve regeneration. 19



ONLINE METHODS Generation of nanoscale and microscale wrinkles. First, polystyrene (PS, glass transition temperature (Tg) = 120 °C) shrinkage film was treated by Ar ion plasma for generating skin layer. By controlling ion plasma power, different skin thickness (h1) was prepared and resulted in different wavelength of nanoscale wrinkle. After plasma treatment, nanoscale wrinkles on the PS substrate were generated by heating at a temperature above the glass transition temperature (i.e., ~120 °C) in the oven (~135 °C) and cooling to room temperature. The areal strain (ε, ε = (A0 - Af)/A0), where A0 and Af denote the areas before and after strain relief, respectively, was controlled by simply changing the heating time in the oven. In here, as plasma power increases from 100 V to 1000 V, wavelength increases from 100 nm to 450 nm at ε = 0.50. For microscale wrinkle, we need to coat polyvinylpyrrolidone (PVP) as a skin layer on the flat PS film. After coating of PVP film, wrinkling process is same with nanoscale wrinkle. After strain relief, PVP layer was removed by ethanol, then underneath PS wrinkle was remained. The feature dimension of the wrinkles was determined from the average measurement of SEM and AFM images with using image processing program (imageJ and XEI program of Park systems). The wavelength is calculated using the Fast Fourier Transform (FFT) using program.

Generation of hierarchical wrinkle. To fabricate multiscale hierarchical wrinkle containing nano- and microscale wrinkles, further process after nanoscale wrinkle is needed. After first generation (G1) of wrinkle, we embedded the G1 wrinkle into polyvinylpyrrolidone (PVP), which served both as a sacrificial skin layer. Then, the PVP-embedded G1 wrinkle was heated 20


Page 20 of 30

Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


at above 130 °C to generate the secondly generated microscale wrinkles (G2 wrinkle). The wrinkle wavelength was modulated by simply controlling the concentration of PVP solution and magnitude of the applied areal strain. After using ethanol to rinse off the PVP, the hierarchical wrinkle system with nano- (G1) and microscale (G2) wrinkles was fabricated over a large area. hNSC culture. Culture of hNSCs derived from the telencephlon (HFT13)23 in undifferentiated state was performed with Dulbecco’s modified Eagle’s medium/Nutrient Mixture F12 (DMEM/F12) medium (Gibco) supplemented with penicillin/streptomycin (2 v/v%, Gibco), N-2 formulation (1 v/v%, Gibco), basic fibroblast growth factor (bFGF) (20 ng/mL, Sigma), leukemia inhibitory factor (LIF) (10 ng/mL, Sigma), and Heparin (8 μg/mL, Sigma). The cells were incubated as neurospheres in humidified air with 5% CO2 at 37 °C. Half of the growth medium was replenished every 2-3 days, and passaging was undertaken every 7-8 days by dissociation of the neurospheres. hNSC differentiation. All the wrinkle substrates were firstly cleaned with 70 % ethanol for 20 min, and washed with D-PBS (Dulbecco’s phosphate-buffered saline) solution for three times. Then the surfaces were coated with mouse natural laminin (10 μg/ml, Invitrogen) by dipcoating for 5 hr. hNSCs dissociated from neurosphere in undifferentiated state was seeded onto the laminin-coated wrinkle structures at a seeding density of 1.5 × 105 cells/ml and maintained under a culture condition without supplementation of bFGF and LIF for 4 days. Immunocytochemistry. hNSCs on each substrate were fixed by incubating with a 4% paraformaldehyde solution (Sigma) for 15 min at room temperature (RT). Fixed cells were incubated with Triton X-100 (0.3 % w/v, Sigma) and BSA (1 % w/v) in D-PBS (Dulbecco’s phosphate-buffered saline) solution for 30 minutes at RT, washed with D-PBS. Samples were 21



then incubated overnight at 4°C with the following primary antibodies: mouse anti-Tubulin β 3 (TUBB3) (1:1000, cat. no. 801201; BioLegend), rabbit anti-glial fibrillary acidic protein (GFAP) (1:1000, cat. no. ab7260; Abcam), and mouse anti-vinculin (1:500, v9131; Sigma). After washing with D-PBS, the resulting samples were stained with goat anti-mouse IgG (H+L) secondary antibody, Alexa fluor 488 (1:250, cat. no. A11001; Invitrogen) or goat anti-rabbit IgG (H+L) secondary antibody, Alexa fluor 633 (1:250, cat. no. A21070; Invitrogen) for 40 min at RT, then with 4′,6-diamidino-2-phenylindole (DAPI, Sigma) for 10 minutes to counterstaining of cell nuclei. Cytoskeleton was stained with Rhodamine Phalloidin (1:40, cat. no. R415; Invitrogen). All fluorescent images were visualized using a confocal laser-scanning microscope (LSM 780, Carl Zeiss). Scanning electron microscopy (SEM). The morphology of wrinkle structures and the cells on adhered onto each substrate were observed by SEM. The cells were fixed with a 4% paraformaldehyde solution (Sigma) for 15 min at room temperature (RT) and washed with DPBS for three times. Then the samples were slightly dehydrated with graded ethanol series ((50%, 70%, 80%, 90%, and 100% for 10 min each) and dried in chamber with silica particles overnight. Then, SEM (S-4800, Hitachi) was used with incident energy of the electron beam between 1 and 10 kV. Cross-sectional images of wrinkle structure was observed by focused ion beam SEM instrument (Helios Nanolab 450 F1, FEI company). Western Blot. hNSCs adhered onto each wrinkle structures were detached by treating TrypLE reagents (Gibco) for 5 min. Total protein was extracted from the detached hNSCs by lysing with RIPA lysis buffer containing a proteinase inhibitor cocktail (ThermoFisher Scientific) on ice for 20 min. Proteins in lysates were quantified through Bradford protein assay kit (Bio-Rad). Equal amounts of protein (15 μg) were loaded to Bolt 4–12% Bis-Tris Plus polyacrylamide gels (ThermoFisher Scientific) for separating by electrophoresis. The separated proteins on gels 22


Page 22 of 30

Page 23 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were then transferred to polyvinylidene difluoride (PVDF) membranes using an iBlot2 transfer system (ThermoFisher Scientific) according to the manufacturer’s instructions. The membranes were blocked by BSA (1%) in Tris-buffered saline (TBS) (Sigma) for 40 min at RT and incubated overnight at 4 °C with following primary antibodies: rabbit anti-phosphoFAK (Tyr39) antibody (1:500, cat. no. 700255; ThermoFisher Scientific), rabbit anti-β-actin antibody (1:1000, cat. no. 4967; Cell Signaling Technology), and mouse anti-vinculin (1:500, v9131; Sigma). Then the samples were treated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:2000, act. no. ab6721; Abcam) or HRP-conjugated goat anti-mouse IgG (1:2000, cat. no. 62-6520; Invitrogen) secondary antibodies for 30 min at RT. The signals from proteins were detected using SuperSignal West Pico Chemiluminescent Substrate (ThermoFisher Scientific) and a ChemiDoc MP system (Bio-Rad). Image analysis. Angle and length of neurites were measured using ImageJ plugin with Tuj1stained images as previously described.32-33 Step-by-step analysis of FA were performed with vinculin-stained fluorescent images combining different plugins of ImageJ programas previously described.34 Actin anisotropy ratio was obtained with FibrilTool, an ImageJ plug-in for quantifying fibrillar structures in raw microscopy images.35 Statistical analysis. Data are expressed as means ± standard error of mean (s.e.m.). Statistical comparisons were conducted using an independent sample t-test or one-way analysis of variance (ANOVA) with Tukey post hoc testing used to make pairwise comparisons between multiple groups. Statistical significance was set to p < 0.05. Data availability. All data supporting the findings of this study are available in the Supplementary Information files.

23



ASSOCIATED CONTENT Supporting Information. Controlled wavelength of wrinkle structures from nano- to microscales by using two different methods for the generation of skin layer. Table containing the wavelengths of nano- and microscale wrinkles. Representative fluorescent images of differentiation markers in hNSCs differentiated on each structure. The SEM images showing cell-to-structure contact. Distribution showing the percentage of neurites having each angle on each structure. Scheme showing scale- and directionality-dependencies of hNSCs on wrinkle structures. Cross-sectional SEM image of hierarchical multiscale wrinkle. Controlled G2 wavelength of hierarchical wrinkle. Number percentage of neurites (θ

Facile fabrication of high-definition hierarchical wrinkle structures for

Recommend Documents