Differentiation and focal adhesion of adipose-derived stem cells on nano-pillars arrays with different spacing

Abstract

Focal adhesion between substrates and cells greatly controls cytoskeletal changes as well as differentiation of the cultivating cells. We here investigated focal adhesion behaviors and differentiation profiles of adipose-derived stem cells on the nano-scaled pillar arrays with diverse pillar-to-pillar distances.

---

Focal adhesion between cells and substrates controls many fundamental cellular behaviors by triggering anchorage-dependent signal pathways related to cell survival, migration, proliferation, and differentiation.^1,2 These focal adhesion points, the primary sites of cell adhesion to the substrate, are composed of various adhesion proteins connecting the substrate to the intracellular cytoskeleton. Thus, through the adhesion complex, the attached cells can recognize the properties of the contacting substrate such as stiffness, nano- or micro-topographical cues and subsequently modulate their cellular behaviors accordingly.^3–6 Cells on the nanoscaled patterns composed of grooves and ridges were shown to preferentially adhere to the ridge rather than channels, suggesting that the focal adhesion sites were linearly organized along the nanopatterned structure, causing elongation and alignment.^7–9 When the pitch-to-pitch distance was increased to microscales, however, the cells were likely to be located on the channel of the micropatterns more than on the ridges and migrate along the channel.¹⁰ Similarly, neuron cells on the nanopillar array showed significantly reduced mobility and were consequently pinned to the nanopillars. They speculated that the top of the vertical pillars acted as pin points for focal adhesion complex. However, the dimension of the respective pillar were shown to be critical for cell anchoring because the narrows pillars tend to penetrate into the cells rather than be attached on the cell surfaces.^11,12

Many researchers recently figured out that the morphological characteristics as well as cell fates are significantly affected by the nano- or micro topographical cues.^13–15 Chen and coworkers demonstrated the important role that cell shape and size can play in directing the fates of human mesenchymal stem cells (hMSCs).¹⁶ hMSCs allowed to adhere, flatten, and spread underwent osteogenesis, while unspread, round cells became adipocytes. Meanwhile, hMSCs were cultured on nanopatterns with gratings of 350 nm linewidth, they were aligned and elongated along the nanogratings and the neuronal gene markers were significantly up-regulated unlike those on unpatterned surface, which indicates that hMSCs shown to transdifferentiate into neuronal-like cells by the induction of morphological changes.¹⁷ Substrate composition and stiffness, applied strain, cell size and spreading have all been shown to influence stem cell fate decisions.^3,18 While adhesion and the degree of cytoskeletal tension are widely accepted as effectors of stem cell fate, the regulatory mechanisms of topographical-stimulated differentiation is not yet established. However, it seems certain that the morphological changes inducing by attachment on the topographical features trigger intracellular signal cascades for differentiation through focal adhesion points. Adipose-derived mesenchymal stem cells (ADSCs) have proven to be a feasible and effective cell source for neural cell replacement therapy because they are similar to marrow stromal cells, and offer advantages such as strong amplification, convenience and lack of immune rejection.^19,20 Thus, in this study, we cultivated ADSCs on the nano-scale pillars on the substrate, which can be recognized as favorable spots for cell adhesion. We hypothesized that the degree of filopodia stretch is dependent on the number of focal adhesion points per respective cell. Morphological changes of the cultivated stem cells were visualized and the genetic profiles were monitored by high-throughput analysis to track down differentiation patterns according to the pillar-to-pillar distance.

In order to manipulate focal adhesion points of ADSCs, the nano-scaled pillar arrays with various pillar-to-pillar distances were lithographed on the PDMS surfaces and the morphology was examined by AFM and SEM (Fig. 1B and C). We here propose that the pillar-to-pillar distance should be carefully optimized so that attached cells on the array can discriminate the difference of the spacing; however, the distance should be within the perimeter of the cellular transformation. Several studies on nano-topography for cell attachments accordingly support our hypothesis; melanoma cells on the dot-shaped PDMS patterns ignored the topographical cues when the spacing of dots was less than 2 μm.²¹ When the spacing between adhesive structures increases over 20 μm, however, cells preferably adhered to the flat surface or one spot on the structure instead of spreading along the adhesive structures.^10,21,22 Thus, we designed the pillars to be distanced by 3 μm to 10 μm and determined the effects of focal adhesion points on cell re-shaping and subsequent trans-differentiation of stem cells. The dimension of the respective nanopillar is also critical to evaluate focal adhesion and the associated cellular differentiation. The height of the respective pillar should be more than 1 μm to maximize the role of pillars as guiding posts for focal adhesion while minimizing undesirable recognition of non-patterned based by the cultivating cell on the pillar array. When the height of the topographical structure is few nanometer, it is hard to precisely control the formation of the focal adhesion points.²³ In addition, soft-lithographed pillars made of PDMS are too elastic and tend to be randomly bent in contact to cells when the aspect ratio is more than 3.²⁴ Thus, we designed the diameter of the pillar to be more than 500 nm to maintain the aspect ratio of 2 and focal adhesion complex can be assembled on the top of the nanopillar. Although soluble factors such as retinoic acid and nerve growth factors have been widely employed for neuronal differentiation of stem cells, the underlying mechanisms of the differentiation induced by chemical cues are considered to be quite different from those by topographical cues, which are not fully elucidated yet.^25–28 Thus, in the current study, the soluble factor-added groups were not considered as positive controls because we only wish to determine the direct effects of the nanotopography. We therefore only employed the PDMS with flat surface as a negative control to examine any effects of the nanopillars on the neuronal differentiation of stem cells.^8,29,30 In addition, the mechanical property of the substrate is important for cellular differentiation. Soft PDMS has been employed as a substrate for stem cell differentiation because the differentiation yield of stem cell is much higher on soft substrates than on rigid substrates.^31–35 Thus, we fixed the volume ratio of curing reagent to base PDMS solution, curing time and temperature to make the stiffness of every PDMS substrate even.

Fig. 1 (A) Schematic diagram of preparing nano-pillar array for controlling cell differentiation. (B) AFM and (C) SEM images of nano-pillar arrays on a PDMS replica. The nano-pillars were 3 μm, 5 μm and 10 μm spacing center-to-center distance with 1 μm height and 500 nm diameter (scale bar = 10 μm).

Morphology of the ADSCs on the nano-pillar arrays were visualized with immunostaining against focal adhesion proteins as shown in Fig. 2 and 3. CLSM reveals that the filopodia of ADSC on the pillar arrays were projected from the body and the lengths of the ‘feet’ were dependent on the pillar-to-pillar distance. Specifically, those on the 5 μm and 10 μm arrays were significantly protruded from the cell body while the 3 μm array showed negligible differences of the protrusion compared to the flat PDMS. The degree of filopodia projection was quantified by imaging-analyzing the confocal images and a cell shape index (CSI) was calculated as well; the CSI result confirmed that ADSCs on the nano-pillar array with 5 μm and 10 μm spacing were more dendritic by 3–4 folds than those on 3 μm spacing or on PDMS surfaces, which suggests that the cell shape became dendritic with increasing the spacing between pillars (Fig. 2B).

Fig. 2 (A) Immunofluorescence micrographs of ADSC cultivated for 12 h on various patterns, showing expression of focal adhesion proteins; F-actin (green), vinculin (yellow), integrin (red) (scale bar = 20 μm). (B) Cell shape index based on the CLSM images of ADSC seeded on the patterns. All data are statistically significant (p < 0.05) by one-way ANOVA (Korea Basic Science Institute).

Fig. 3 (A) Immunofluorescence for pFAK (green) and F-actin (red) on filopodia of ADSC cultivated for 6 h on various patterns with 3 μm, 5 μm, and 10 μm spacing. The inset is an enlargement of filopodia adhered with nano-pillars. (B) Z-stack images of focal adhesion point at the filopodia (a) and cell body (b) on the nano-pillar array with 3 μm spacing (scale bar = 50 μm) (Central Lab Kangwon National University).

Furthermore, ADSCs seems to employ each pillar unit as a contact point for cell adhesion because the focal adhesion proteins such as vinculin, integrin, and pFAK were mainly distributed along the top of each pillar (Fig. 2A and 3A). We observed that the filopodia of ADSCs on the nano-pillars grabbed the pillars and the lamellipodia widely spread on the PDMS with flat surface. Additionally, pFAK was highly localized at the filopodia and deeply anchored to the spot where the nano-pillars support the adhesion of filopodia while flat cell body was on the top of the pillars (Fig. 3B). Notably, 5 μm and 10 μm spacing showed higher localization as well as the number of localized points of the pFAK around pillars than 3 μm spacing. However, those points are too small to be clearly visualized in Z-stacked images. Thus, we presented the Z-stack images of 3 μm spacing (Fig. 3B), which showed the CLSM-observable sized focal adhesion point of pFAK.

We expected that the gene expression of ADSCs on the nano-pillars was influenced by morphological change due to continuous interactions with the pillars during proliferation. Fig. 2 and 3 suggest that the nano-pillars affect to cell adhesion as well as morphology by creating adhesion junctions with filopodia during cell migration and proliferation. Here, we suggest that the dimension of each pillar is recognizable by ADSCs and the each contact points between cells and the pillars can recruit focal adhesion proteins to stimulate intracellular signals for modulate cell shapes. To analyze the expression levels of specific markers related to focal adhesion points, we further performed microarray. The several genes related focal adhesion points including integrin β1 (ITGB1), Rho guanine nucleotide exchange factors (ARGHEF5), and ADAM-like protein 4 (ADAMTSL4) were highly activated in ADSCs cultivated on the nanopillar array (Fig. S2†). Due to the short incubation period of 5 days, the expression levels of specific markers should be determined at the mRNA levels, not protein levels. ADSCs cultivated on the nanopillar array for 5 days are not enough to be fully differentiated because the differentiation yield of stem cells is usually low when differentiation is induced only by topographical cues, without differential media including soluble factors.^17,36 In addition, based on the cell viability result, ADSCs on the nanopillars showed similar proliferation behaviors to those on flat PDMS, suggesting that pillar structure and spacing affected morphological change and subsequent gene expression of ADSCs (Fig. S5†). Numerous studies indicated that the morphological change of stem cell by topographical cues can be subject to differentiation as well as cell migration and proliferation.^37,38 Especially, stem cells cultivated on the aligned micro- or nano-patterns were shown to experience morphological changes such as alignments and elongation and such dendritic changes could cause neuronal differentiation.^17,39 Therefore, we employed a high-throughput method to track down the gene expression patterns of ADSCs on the nano-pillar arrays with various spacing compared to that on the flat surface. When the microarray analysis was performed for the RNA extract from the cultivated ADSCs on each sample, the correlation plots were obtained for the whole genes of ADSCs on the nanopillar with those on the flat surface as a control (Fig. 4A). The analysis of Euclidean distance metric selection also revealed that 5 μm and 10 μm spacing are clustered closely but separately from 3 μm spacing. The correlation plots based on the clustering are presented in the average signal intensity (x-axis) vs. M-value (y-axis), a normalized signal intensity of the respective gene in comparison to PDMS surfaces. Thus, the plot suggests significant information which sample changed the gene expression pattern of the cultivated cells by up-regulation (red dot) or down-regulation (green dot) and the extent of deviation from M-value of 1 indicates the change of the gene expression levels.

Fig. 4 Microarray results of mRNA from cell on the PDMS pillar arrays with respect to flat PDMS surfaces. (A) Correlation plots of M-values between cells cultivated on the pillar arrays (3 μm, 5 μm, or 10 μm spacing) and those on flat PDMS. (B) Heat maps and hierarchical clustering of significant genes filtered by neurogenic differentiation. (C) The number of significant gene related to total cell differentiation and neurogenic differentiation. (D) Gene fold expression of neurogenic and cytoskeletal markers.

When the spacing between the nano-pillars increased from 3 μm to 10 μm, numerous gene expressions were remarkably up-regulated in comparison to the flat PDMS (Fig. 4B). Specifically, the number of the significant genes related to differentiation was 67, 206, and 239 for ADSCs on the nano-pillar arrays with 3, 5 and 10 μm spacing, respectively (Fig. 4C). Specifically, the majority of the significant genes related differentiation was strongly associated with neuronal differentiation; 3 μm spacing (34.4%), 5 μm spacing (47.9%), and 10 μm spacing (44.8%) (Fig. 4C). For example, the number of the gene associated with the neuronal differentiation in 10 μm spacing was 101, whereas that with the osteogenic, chondrogenic and adipogenic differentiation was 21, 14, and 8, respectively (Fig. S3†). Upon increasing the pillar-to-pillar distance, the large parts of the neural differentiation genes were up-regulated; the number of the up-regulated neuronal differentiation gene were 2.4 folds and 4.1 folds higher compared to those down-regulated for the cells on 5 and 10 μm spacing patterns, respectively. Thus, these results suggest that the neuronal differentiation was significantly provoked by the nano-pillar structure with increasing the distance between pillars, especially over 5 μm spacing. In addition, we confirmed that the major markers associated with the neuronal differentiation such as nestin, glial cell derived neurotrophic factor (GFAP), tyrosine hydroxylase (TH), SOX1, SOX2 and neurofilament (NEF) were highly up-regulated in the pillar array compared to those in flat PDMS (Fig. 4D). However, β-tubulin III or microtubule associated protein 2 (MAP2), which are mature neuronal makers, showed no significant expression change, suggesting that neuronal differentiation was still progressing post-cultivation on the nano-pillars at day 5. We observed that the nanopillar arrays were fully covered by the ECM due to the proliferation of ADSCs after 5 days cultivation, which indicated that the unique topography of the surface was lost (Fig. S4†). The topographical cues are likely control the differentiation of the cultivating cells at the early stage of the incubation period or at the newly proliferating cells on the topographical cues. Thus, we here focused on what happens during the initial stages of the cellular differentiation on the nanopillar arrays, not for a long term period. In addition, actin filaments, actin-related proteins and phosphatase were up-regulated on the nano-pillars, which suggesting that alteration of cytoskeleton rearrangement by the nano-pillar structure could lead to a change in gene regulation for cell differentiation. Several studies indicated that induction of cytoskeleton rearrangement would play an important role in creating the signal transduction for the transdifferentiation.^13,17,40,41 Based on the results in the current study, it would be clear that dendritic elongation of cell shape induced by pillars, localization of focal adhesion complex at filopodia, and active gene expression involved in neuronal differentiation are not individual phenomena and a coincidence. Thus, we firmly believe that the filopodia (dendritic cellular) extension to find the nano-pillars induced an extensive actin stretching, which triggered cytoskeleton rearrangement and newly created intrinsic signal transduction for neuronal differentiation. In addition, we also demonstrated that nano-pillars played an important role in regulating stem cell differentiation without the presence of biochemical signals.

Conclusions

Nano-scale pillar array with the pillar-to-pillar distances ranging from 3 μm to 10 μm was fabricated on the PDMS. The degree of filopodial protrusion of ADSCs cultivated on the nano-pillar arrays significantly increased when pillar-to-pillar distance increased from 3 μm to 10 μm, which followed by dendritic morphological changes. In addition, the focal adhesion proteins were highly localized on the respective pillar. Furthermore, the morphological changes of ADSCs by nano-pillar arrays affected the gene expression of ADSCs and specially, most of the significant genes associated with neuronal differentiation were strongly up-regulated with increasing the pillar-to-pillar distance. Thus, the nano-pillar arrays with 5 μm and 10 μm spacing induced the filopodia extension of ADSCs, which was accompanied by an extensive actin stretching and cytoskeleton rearrangement, and eventually the morphological changes newly created intrinsic signal transduction for neuronal differentiation.

Acknowledgements

This work was supported by a grant from the Ministry of Health and Welfare of Republic of Korea (HI12C0581) and Kangwon National University.

Notes and references

1. A. De Arcangelis and E. Georges-Labouesse, Trends Genet., 2000, 16, 389.
2. E. H. Danen and A. Sonnenberg, J. Pathol., 2003, 201, 632.
3. J. S. Park, J. S. Chu, A. D. Tsou, R. Diop, Z. Tang, A. Wang and S. Li, Biomaterials, 2011, 32, 3921.
4. M. Théry, J. Cell Sci., 2010, 123, 4201–4213.
5. C. H. Seo, H. Jeong, K. S. Furukawa, Y. Suzuki and T. Ushida, Biomaterials, 2013, 34, 1764.
6. Y. Wang, G. Wang, X. Luo, J. Qiu and C. Tang, Burns, 2012, 38, 414.
7. D. H. Kim, E. A. Lipke, P. Kim, R. Cheong, S. Thompson, M. Delannoy, K. Suh, L. Tung and A. Levchenko, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 565.
8. E. K. Yim, R. M. Reano, S. W. Pang, A. F. Yee, C. S. Chen and K. W. Leong, Biomaterials, 2005, 26, 5405.
9. F. C. M. J. M. van Delft, F. C. van den Heuvel, W. A. Loesberg, J. te Riet, P. Schön and C. G. Figdor, Microelectron. Eng., 2008, 85, 1362.
10. S. Yoon, Y. K. Kim, E. D. Han, Y. Seo, B. H. Kim and M. R. K. Mofrad, Lab Chip, 2012, 12, 2391.
11. C. Xie, L. Hanson, W. Xie, Z. Lin, B. Cui and Y. Cui, Nano Lett., 2010, 10, 4020.
12. A. K. Shalek, J. T. Robinson, E. S. Karp, J. S. Lee, D. R. Ahn, M. H. Yoon, A. Sutton, M. Jorgolli, R. Gertner, T. Gujral, G. MacBeath, E. G. Yang and H. Park, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 1870.
13. K. A. Kilian, B. Bugarija, B. T. Lahn and M. Mrksich, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 4872.
14. J. T. Connelly, J. E. Gautrot, B. Trappmann, D. W. Tan, G. Donati, W. T. Huck and F. M. Watt, Nat. Cell Biol., 2010, 12, 711.
15. K. Kolind, K. W. Leong, F. Besenbacher and M. Foss, Biomaterials, 2012, 33, 6626.
16. R. McBeath, D. M. Pirone, C. M. Nelson, K. Bhadriraju and C. S. Chen, Dev. Cell, 2004, 6, 483.
17. E. K. Yim, S. W. Pang and K. W. Leong, Exp. Cell Res., 2007, 313, 1820.
18. L. E. McNamara, R. J. McMurray, M. J. Biggs, F. Kantawong, R. O. Oreffo and M. J. Dalby, J. Tissue Eng., 2010, 2010, 120623.
19. J. M. Gimble, A. J. Katz and B. A. Bunnell, Circ. Res., 2007, 100, 1249.
20. D. A. De Ugarte, K. Morizono, A. Elbarbary, Z. Alfonso, P. A. Zuk, M. Zhu, J. L. Dragoo, P. Ashjian, B. Thomas, P. Benhaim, I. Chen, J. Fraser and M. H. Hedrick, Cells Tissues Organs, 2003, 174, 101.
21. D. Lehnert, B. Wehrle-Haller, C. David, U. Weiland, C. Ballestrem, B. A. Imhof and M. Bastmeyer, J. Cell Sci., 2004, 117, 41.
22. K. Sheets, S. Wunsch, C. Ng and A. S. Nain, Acta Biomater., 2013, 9, 7169.
23. J. M. Łopacińska, C. Grădinaru, R. Wierzbicki, C. Købler, M. S. Schmidt, M. T. Madsen, M. Skolimowski, M. Dufva, H. Flyvbjerg and K. Mølhave, Nanoscale, 2012, 4, 3739.
24. J. L. Tan, J. Tien, D. M. Pirone, D. S. Gray, K. Bhadriraju and C. S. Chen, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 1484.
25. Y. Okada, T. Shimazaki, G. Sobue and H. Okano, Dev. Biol., 2004, 275, 124.
26. P. D. Tonge and P. W. Andrews, Differentiation, 2010, 80, 20.
27. B. Inanç, A. E. Elçin and Y. M. Elçin, Tissue Eng., Part A, 2008, 14, 955.
28. L. Calzà, A. Giuliani, M. Fernandez, S. Pirondi, G. D'Intino, L. Aloe and L. Giardino, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 7325.
29. A. Béduer, C. Vieu, F. Arnauduc, J. Sol, I. Loubinoux and L. Vaysse, Biomaterials, 2012, 33, 504.
30. E. Kingham, K. White, N. Gadegaard, M. J. Dalby and R. O. C. Oreffo, Small, 2013, 9, 2140.
31. J. Fu, Y. Wang, M. T. Yang, R. A. Desai, X. Yu, Z. Liu and C. S. Chen, Nat. Methods, 2010, 7, 733.
32. F. Chowdhury, Y. Li, Y. Poh, T. Yokohama-Tamaki, N. Wang and T. S. Tanaka, PLoS One, 2010, 5, e15655.
33. S. S. Kumar, J. H. Hsiao, Q. D. Ling, I. Dulinska-Molak, G. Chen, Y. Chang, Y. Chang, Y. H. Chen, D. C. Chen, S. T. Hsu and A. Higuchi, Biomaterials, 2013, 34, 7632.
34. S. Nemir and J. L. West, Ann. Biomed. Eng., 2010, 38, 2.
35. W. L. Murphy, T. C. McDevitt and A. J. Engler, Nat. Mater., 2014, 13, 547.
36. E. Engel, E. Martinez, C. A. Mills, M. Funes, J. A. Planell and J. Samitier, Ann. Anat., 2009, 191, 136.
37. K. S. Brammer, C. Choi, C. J. Frandsen, S. Oh and S. Jin, Acta Biomater., 2011, 7, 683.
38. X. Wang, W. Song, N. Kawazoe and G. Chen, Soft Matter, 2013, 9, 4160.
39. Y. I. Cho, J. S. Choi, S. Y. Jeong and H. S. Yoo, Acta Biomater., 2010, 6, 4725.
40. J. N. Lavoie, M. C. Landry, R. L. Faure and C. Champagne, Cell. Signal., 2010, 22, 1604.
41. W. Song, N. Kawazoe and G. Chen, J. Nanomater., 2011, 2011, 265251.

---

Footnote

† Electronic supplementary information (ESI) available: SEM images of the nano-holes on a silicon wafer (Fig. S1), microarray results of focal adhesion points and differentiation (Fig. S2 and S3), SEM image of ADSCs on the pillar array for 7 days (Fig. S4) and MTT assay (Fig. S5). See DOI: 10.1039/c5ra07608k

---