Thickness sensing of hMSCs on collagen gel directs stem cell fate
Wen Shing Leong,Chor Yong Tay,Haiyang Yu,Ang Li,Shu Cheng Wu,Duong-Hong Duc,Chwee Teck Lim,Lay Poh Tan
DOI: https://doi.org/10.1016/j.bbrc.2010.09.052
IF: 3.1
2010-01-01
Biochemical and Biophysical Research Communications
Abstract:Research highlights ► hMSCs appeared to sense thin collagen gel (130 μm) with higher effective modulus as compared to thick gel (1440 μm). ► Control of collagen gel thickness can modulate cellular behavior, even stem cell fate (neuronal vs. Quiescent). ► Distinct cellular behavior of hMSCs on thin and thick collagen gel suggests long range interaction of hMSCs with collagen gel. Abstract Mechanically compliant substrate provides crucial biomechanical cues for multipotent stem cells to regulate cellular fates such as differentiation, proliferation and maintenance of their phenotype. Effective modulus of which cells sense is not only determined by intrinsic mechanical properties of the substrate, but also the thickness of substrate. From our study, it was found that interference from underlying rigid support at hundreds of microns away could induce significant cellular response. Human mesenchymal stem cells (hMSCs) were cultured on compliant biological gel, collagen type I, of different thickness but identical ECM composition and local stiffness. The cells sensed the thin gel (130 μm) as having a higher effective modulus than the thick gel (1440 μm) and this was reflected in their changes in morphology, actin fibers structure, proliferation and tissue specific gene expression. Commitment into neuronal lineage was observed on the thin gel only. Conversely, the thick gel (1440 μm) was found to act like a substrate with lower effective modulus that inhibited actin fiber polymerization. Stem cells on the thick substrate did not express tissue specific genes and remained at their quiescent state. This study highlighted the need to consider not only the local modulus but also the thickness of biopolymer gel coating during modulation of cellular responses. Keywords hMSCs Collagen Neurongenesis Quiescent Thickness 1 Introduction There are increasing evidences in recent years that living cells are responsive to environmental signals not only of chemical but also physical nature [1–7] . One of these physical cues that have been gaining attention lately is the substrate elasticity [8,9] . Substrate elasticity has been shown to affect cellular behavior including morphology, survival, proliferation, locomotion and differentiation [10–16] . Cells probe the substrate elasticity by exerting traction force at the anchorage points (focal adhesion), and sense the generated reactive force from the substrate [3,17] . Soft substrate generate small reactive force and hence induce lower cytoskeleton tension and lesser polymerization in stress fiber [17] . As a result cells appear to be round and small. By contrast, stiff substrates with higher reactive forces support formation of relatively rigid, stiff and contractile stress fibers accompanied by a well spread cell morphology. Stem cell is a promising tool for tissue therapy, especially for degenerative disease like myocardial infarction, Parkinson’s syndrome, etc. Optimal differentiation of stem cell into specific tissue line ages occur only on substrate with modulus similar to the native tissues, which is lower than 30 kPa for soft tissues such as neurons and myoblast [8,18] . With even softer modulus, 250 Pa, hMSCs become quiescent while preserving their multilineage potential [19] . Apart from the intrinsic substrate modulus, thickness of the substrate has also been shown to determine the effective mechanical properties of which cells will sense. Ability of cells to remodel their environment enable them to sense not only the local (immediate at point of contact) but also global mechanical properties (deformed region at a distance away from point of contact) [3] . Interaction depth of cells with compliant synthetic gel such as Polyacrylamide (PA) gel has been simulated and demonstrated in the range of a few to several tens of microns [20–23] . Within the interaction depth, the effective substrate modulus of which cells sense is higher than the intrinsic modulus due to the interference from rigid support beneath the substrate. As a result cells may behave differently from what is expected. However, unlike synthetic gel (e.g. elastomeric PA) that has perfect flexible chain, many common biological polymers (e.g. fibrin, collagen, actin cytoskeleton, etc.) are semi-flexible and undergo strain hardening effect. The elastic modulus of these materials increases as being strained beyond a certain extension [24,25] . Recent study using fibrin gel (4 mg/ml) of shear modulus 120 Pa has demonstrated large lateral interaction distance of hMSCs, which can be up to 450 μm away from the center of cell [26] . The reported distance is much larger than lateral interaction distance of 10 μm as suggested on PA gel [20] . Therefore, synthetic gel might not be representative in predicting cellular response in biological matrices. Even though large lateral interaction distance of hMSCs on natural polymer gel has been demonstrated using fibrin, the vertical interaction depth and effect of thickness sensing on hMSCs behavior remains unknown. Among biological polymers, type I collagen represents extracellular matrix (ECM) the best as it is the most prevalent protein in the body [27] . Therefore, collagen gel has been widely used for cell–ECM interaction study [16,28,29] . Here we report the use of type I collagen as the substrate to modulate hMSCs behavior by varying the substrate thickness. Using collagen gel (3 mg/ml) of similar shear modulus to the reported fibrin gel, we demonstrated that the interaction depth of hMSCs could be up to hundreds of micron and distinct stem cell fate could be modulated without changing the mechanical properties but just the thickness of collagen gel. 2 Materials and methods 2.1 Human mesencymal stem cells (hMSCs) culture Bone marrow-derived hMSCs (Cambrex, USA) (lot number: 6F4393 & 8F3211) were expanded and maintained in MSCGM (Cambrex, USA) prior to experiment. Experiments were carried out using hMSCs between Passages 4 and 8 and cultured in low glucose DMEM supplemented with 10% Gold FBS, 1% l -glutamine and 1% penicillin–streptomycin (all culture components from PAA Laboratories, Pasching, Austria). hMSCs were seeded at 1500 cells/cm 2 onto thin (130 μm), thick (1440 μm) collagen gel and collagen molecules coated coverglass in 24-wellplate for 2 weeks. Adipogenic and osteogenic differentiation were induced in adipogenic and osteogenic induction medium (Invitrogen, USA) for 14 and 21 days, respectively, for multipotency study. 2.2 Collagen gel preparation Collagen gel was covalently bonded onto glutaraldehyde (GA) grafted aminosilanized coverglass. Briefly, cleaned 13 mm diameter coverglass was immersed in 5 v/v% aminopropyltriethoxysilane (APTES) for 5 min, followed by rinsing three times with acetone and then cured at 110 °C overnight. It was then immersed in 2.5% GA for 3 h and rinsed thoroughly with PBS. Rat tail type I collagen (Sigma–Aldrich, USA) was dissolved in 0.5 M acetic acid to obtain a stock solution with a concentration of 7.142 mg/ml. Gelation of 3 mg/ml collagen was done through the neutralization process. Neutralized solution (0.35 and 0.05 ml) was spread onto the GA-grafted aminosilanized coverglass in 24-wellplate to obtain collagen gel with thicknesses of 1440 and 130 μm, respectively. Finally, the collagen gel was obtained after incubation at 37 °C overnight. Needle was traced along the boundary of collagen gel to eliminate adhesion to the wall of wellplate prior to cell seeding. Collagen molecules coated coverglass was obtained by immersing GA-grafted aminosilanized coverglass in 3 mg/ml collagen solution overnight followed by three times washing with 1× PBS. 2.3 Collagen gel characterization See Supplementary materials . 2.4 Tissue specific gene expression profile Total RNA of hMSCs was isolated after 2 and 14 days of culture on different groups of substrates using MN NucleoSpin RNA II Kit (Macherey–Nagel, Germany). Quantitative real time RT-PCR was studied using CFX96 real time PCR detection system (Biorad., USA). House keeping and tissue specific genes were selected from primer bank ( http://pga.mgh.harvard.edu/primerbank/ ) as listed in Table 1 . For details see Supplementary materials . 2.5 Immunocytochemistry hMSCs on different groups of substrates were fixed in 4% paraformaldehyde (Sigma–Aldrich, USA) for 15 min and permeabilized in 0.1% Triton-X for 10 min after 7 and 14 days of culture. Nonspecific binding sites were blocked using 5% BSA solution (Invitrogen, USA). Intracellular actin filaments were stained with Rhodamine conjugated Phalloidin (Chemicon, Millipore, USA) while collagen molecules were stained with Anti-type I collagen (Sigma–Aldrich, USA), and TRITC anti-mouse IgG (Invitrogen, USA). Samples were imaged by Nikon Eclipse 80i Fluorescent Microscopy (Nikon, Japan) with 20× Plan Fluor (N.A. 0.55) and 40× Plan Fluor (N.A. 0.75) objective lens. 2.6 Histochemistry See Supplementary materials . 2.7 Cell morphology imaging and characterization Phase contrast micrographs of cells at a day after cell seeding were taken using 10× objective lens with an inverted light microscope, Nikon Eclipse TS100 (Nikon, Japan). Cells area was measured using Image J, ( http://rsb.info.nih.gov.ij/ ). For each data point more than 20 cells were measured from three different substrates. 2.8 BrdU incorporation assay Cells were plated on substrates for 48 h and 2 weeks before being incubated in complete medium containing BrdU labeling reagent (Zymed, Invitrogen, USA) overnight. Cells were stained with BrdU staining kit and cell nucleus was counter stained with Hematocylin. Data are expressed as means ± standard errors of three independent experiments with more than 300 cells counted. 2.9 Statistics P -values were calculated using unpaired Student’s t -test; p < 0.05 were considered significant. 3 Results 3.1 Collagen substrates of different thickness have consistent localized Young’s modulus The thick and thin collagen gels on coverglass had thickness of 1446 ± 58.81 and 134.83 ± 15.04 μm, respectively, as measured by Raman Confocal Scanning ( Fig. S1 ). To ensure consistent surface chemistry, collagen molecule coated coverglass was used as control group. Based on literature, the thickness of a monolayer collagen coating would be around 1–1.5 nm [30] . The presence of collagen molecules was confirmed by the positive staining of anti-type I collagen ( Fig. S2 ). To estimate the initial modulus sensed by hMSCs upon adhesion, microelasticity of different groups of substrates were measured using AFM. As expected, both thick and thin gels have similar localized Young’s modulus of 2 kPa as shown in Fig. S3 . These results confirm that the two sets of collagen substrate used have consistent local material mechanical properties and that gel thickness is the only variable for subsequent studies. 3.2 Cellular spreading area and morphology is modulated by thickness of collagen gel Culturing hMSCs on collagen substrate of differing thickness resulted in drastic alteration in terms of cellular morphologies. After 2 days in vitro culture, highest degree of cell spreading was observed for the control group with mean projected area of ∼7500 μm 2 . Cellular spreading area was found to decrease on collagen gel, which is in accordance with previous studies involving the use of substrate of varying elastic modulus showing that cells cultured on softer matrix (lower Young’s modulus) exhibited a lower degree of spreading [9,31] . Interestingly, as in our system where the Young’s modulus of collagen substrates was constant, the projected cell area was found to decrease with increasing gel thickness. The projected cellular area decreased by approximately 1.5-fold from 3907 to 2751 μm 2 as the gel thickness increased from 130 to 1400 μm ( Fig. 1 A). Besides, hMSCs cultured on thin collagen gel exhibited neurite-like extensions ( Fig. 1 B). This morphology resembles hMSCs that have undergone neurogenesis [8] . 3.3 Thick collagen gel suppresses actin fiber polymerization, cellular proliferation and tissue specific gene expression As cellular cytoskeletal network has previously been shown to play a critical role in directing stem cell fate [32,33] . The immuno-labeled cellular actin filament to visualize the cytoskeletal development and rearrangement of cells cultured on collagen gel of differing thickness. As shown in Fig. 2 C and F, cells cultured on the control substrate exhibited well defined microfilaments. Bundles of microfilaments appeared to be thick and aligned in parallel with the cells major axis of spreading. Conversely, integrity of actin cytoskeleton bundles decreased in the presence of collagen gel. Thinner bundles of actin-cytoskeleton were found within hMSCs seeded on thin collagen gel (130 μm) as compared to control group ( Fig. 2 B and E). Even though the cells on thin collagen gel had similar long microfilaments, there was visually a decrease in fiber bundles’ integrity and connectivity, together with increase in short actin fiber bundles. Smallest actin bundle size was found on thick collagen gel (1440 μm). hMSCs cultured on thick collagen gel gave rise to the most disorganized microfilaments as observed from day seven onwards ( Fig. 2 A and D). Unlike thin collagen gel and control group, thick collagen gel did not support formation of long and integrated microfilaments. Instead, associated actin fibers were so sparse that they could hardly be defined and appeared to be short, branched and scattered as viewed under microscope. Dotted rhodamine–phalloidin became obvious with culturing time. Absence of integrated microfilaments suggests that thick collagen gel is not as supportive as thin collagen gel for filamentous actin polymerization in hMSCs. To investigate if gel thickness plays a role in regulating cellular proliferation rate, BrdU incorporation assay was carried out with cells cultured on substrates of different thickness at 2nd and 14th day of culturing. None of the cells on thick gel were incorporated with BrdU at both timepoints studied. Lack of BrdU incorporation indicated suppression of DNA synthesis on thick gel, a required step for cell division. On the other hand, cells on thin gel demonstrated a lower proliferative rate than control group as evidenced by lesser number of BrdU incorporated cells ( Fig. S4 ). In addition to the absence of proliferation activity, cells cultured on the thick collagen gel demonstrated downregulation in mRNA expression level of all studied tissue specific genes as normalized to control group ( Fig. 3 A and B). With similar cDNA concentration input during qPCR amplification, control group had amplification threshold of tissue specific mRNA between 25 and 34 cycles associated with specific melting temperature while thick collagen gel did not show up any amplification within 40 cycles of PCR amplification studied. Neither neurogenic nor osteogenic specific marker was expressed in hMSCs cultured on thick collagen gel as analyzed at 2nd and 14th day of culture. Downregulation of tissue specific genes and non proliferative behavior suggest that stem cells culture on the thick collagen gel remain quiescent. Multilineage potential of the suggested quiescent cell has been confirmed by induced differentiation of hMSCs into adipocytes (on thick collagen gel) and osteocytes (upon trypsinized from thick collagen gel and plated onto wellplate) after 2 days of culturing on thick collagen gel (see Supplementary material ). 3.4 Thin collagen gel promotes neurogenesis mRNA level of neuron- and bone-specific markers in hMSCs was normalized to control group after 2nd and 14th days of culturing on thin collagen gel ( Fig. 3 A and B). After 2 days of culturing, neurogenic glial mature marker (GFAP) was upregulated 1.38-fold while osteogenic early marker, ALPL, was downregulated five-fold. No significant difference was observed for the rest of the markers studied. With further culturing the cells up to 14 days, tissue-specific commitment was observed. The cells showed significant upregulation in all neurogenic markers analyzed (neurogenic transcription factor, NeuroD1, neuronal mature marker, MAP2, and GFAP), but downregulation in osteogenic transcriptional factor, RUNX2, and ALPL. No significant difference in osteogenic differentiation marker, Osteonectin, was observed. Therefore the changes in mRNA level suggest the possibility to direct commitment of hMSCs into neuronal lineage on thin collagen gel. This is in contrast to the thick substrate where upregulation of neuronal markers were not observed in both 2nd and 14th day of culture. Commitment of hMSCs on thin gel into neurogenic lineage is further confirmed by expression of early and mature neuronal protein markers, Nestin and MAP2, respectively. Nestin and MAP2 are positively stained in hMSCs cultured on thin collagen gel for 14 days but not in control group ( Fig. 4 ). Negative staining of Alizarin red S in the cells cultured on thin collagen ( Fig. S6 ) again suggests specific differentiation into neuronal lineage of those cells. 4 Discussion 4.1 hMSCs detect thick gel as softer substrate than thin gel Cells on soft substrate has been reported to exhibit small cell area and less-developed actin stress fiber due to the lack of reactive forces upon cell probing [9] . In this study, cell area and development of cytoskeleton structure of hMSCs decreased significantly as the collagen gel thickness increased from 130 to 1440 μm. All these cellular behaviors suggest that thick gel has too small a reactive force to trigger the actin fiber formation and cell spreading just like the effect of a soft substrate. In other words, the cells sensed the 1440 μm thick collagen gel as a “softer” material as compared to thin collagen gel even though the local modulus as measured by AFM is similar. 4.2 hMSCs remain at quiescent state on thick gel Thick collagen gel halted the cell growth cycle as confirmed by BrdU assay. These non proliferative cells remained at their quiescent state as evidenced by their suppressed tissue specific gene expression. Additionally, the capability to differentiate into adipocytes and osteocytes was proven by induction studies ( Fig. S5; Supplementary materials ). Furthermore, the cells appear to have smaller cell size and disorganized F-actin. This quiescent cell state is similar to the phenomenon of culturing hMSCs on very soft substrate of about 250 Pa in shear modulus [19] . Shear modulus of 3 mg/ml rat tail type I collagen that were employed in this study has also been reported in the similar range, ∼150 Pa [34,35] . Therefore, hMSCs possibly detected effective modulus which is closer to bulk modulus on thick collagen gel. 4.3 Interaction depth of hMSCs on collagen gel between 130 and 1440 μm Distinctively different cellular behaviors, including cell morphology, cytoskeleton arrangement, cell size, proliferation rate and specific tissue lineage commitment, have been observed in hMSCs upon culturing on collgen gel of different thickness, i.e. of millimeter-, thick gel (1440 μm), and micron-, thin gel (130 μm), range. Such observation demonstrated the ability of hMSCs to recognize the difference of the two thicknesses, in which thin gel was interpreted with “higher effective modulus” than thick gel despite their identical local modulus. hMSCs likely probing the mechanical properties of thin gel with interference from underlying coversilp at hundreds of microns away upon cell adhesion. Consequently, the higher effective mechanical properties of thin gel induced specific cellular activities as reported in this study. Therefore, interaction depth of hMSCs is not only restricted at several to tens of micron [20–22] which is commonly reported on PA gel, but can be up to hundreds of microns as demonstrated in our study using collagen gel. The deviation of our findings from reported cell sensing depth is likely due to the difference in substrate studied. Unlike synthetic gels, biopolymers display nonlinear strain stiffening properties [25] . In the case of collagen gel, the critical strain is 5% [34] but cell induced-deformation in collagen gel could be up to 20% [36] , which is obviously in the strain stiffening regime. Recently, it has been reported that hMSCs can interact laterally up to 450 μm on fibrin gel of shear modulus 140 Pa [26] . Winer et al. demonstrated that the mechanosensor of mesenchymal cell type is stress but not strain limited. Therefore, cells apply traction force and respond to the generated resistance from substrate which forms a feedback loop until they reach equilibrium. Unlike synthetic polymers, biopolymers of similar modulus will generate higher resistance upon cell probing due to their strain stiffening properties which in turn induce larger cell traction forces. As a result, higher cell traction force is generated on biopolymer which allows further interaction distance. Therefore, long range interaction of hMSCs on collagen gel allows interference from underlying coverglass on thin but not thick gel. As a result, thin gel gives rise to higher resistance upon cell contraction and exhibits higher effective modulus that supports series of cell activities. However, for thick gel of thousands of microns, there is no interference from underlying coverglass. Therefore the intrinsic bulk modulus of collagen gel becomes the dominant factor. As a result, the thick gel demonstrated lower effective modulus which inhibits a series of cell activities. 5 Conclusions This work showed that substrate thickness apart from intrinsic modulus plays an important part in mechanosensing of hMSCs. With the same local Young’s modulus, thin collagen gel of 130 μm promoted actin cytoskeleton polymerization and neuronal lineage commitment of hMSCs. However, thick collagen gel of 1440 μm induced non-developed cytoskeleton structures, suppressed cellular proliferation and maintained the quiescent cell state of hMSCs where no tissue specific genes were expressed throughout the culture period studied. The results suggest that thick collagen gel has a smaller effective modulus as sensed by hMSCs. We have demonstrated modulation of hMSCs behavior via changing collagen thickness for the first time. hMSCs is expected to have interaction depth between 130 and 1440 μm with collagen gel. The findings revealed the importance of considering substrate thicknesses when designing collagen based substrates especially for soft tissue engineering. 6 Conflict of interest None of the authors have a conflict of interest associated with this work. Acknowledgments The authors acknowledge A∗Star, Singapore ( SERC Grant No: 072 101 0021 ) and NTU for financial support. The authors thank National Heart Centre for the valuable discussions. Appendix A Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbrc.2010.09.052 . Appendix A Supplementary data Supplementary materials Supplementary materials. References [1] B. Geiger J.P. Spatz A.D. Bershadsky Environmental sensing through focal adhesions Nat. Rev. Mol. Cell Biol. 10 2009 21 33 [2] N. Wang J.D. Tytell D.E. Ingber Mechanotransduction at a distance. mechanically coupling the extracellular matrix with the nucleus Nat. Rev. Mol. Cell Biol. 10 2009 75 82 [3] V. Vogel M. Sheetz Local force and geometry sensing regulate cell functions Nat. Rev. Mol. Cell Biol. 7 2006 265 275 [4] M.J. Dalby N. Gadegaard R. Tare A. Andar M.O. Riehle P. Herzyk C.D.W. Wilkinson R.O.C. Oreffo The control of human mesenchymal cell differentiation using nanoscale symmetry and disorder Nat. Matter 6 2007 997 1003 [5] C.Y. Tay H. Gu W.S. Leong H. Yu H.Q. Li B.C. Heng H. Tantang S.C.J. Loo L.J. Li L.P. Tan Cellular behavior of human mesenchymal stem cells cultured on single-walled carbon nanotube film Carbon 48 2010 1095 1104 [6] C.Y. Tay H. Yu M. Pal W.S. Leong N.S. Tan K.W. Ng D.T. Leong L.P. Tan Micropatterned matrix directs differentiation of human mesenchymal stem cells towards myocardial lineage Exp. Cell Res. 316 2010 1159 1168 [7] P.A. Janmey J.P. Winer M.E. Murray Q. Wen The hard life of soft cells Cell Motil. Cytoskeleton 66 2009 597 605 [8] A.J. Engler S. Sen H.L. Sweeney D.E. Discher Matrix elasticity directs stem cell lineage specification Cell 126 2006 677 689 [9] D.E. Discher P. Janmey Y.L. Wang Tissue cells feel and respond to the stiffness of their substrate Science 310 2005 1139 1143 [10] G.W. Rebecca The role of matrix stiffness in regulating cell behavior Hepatology 47 2008 1394 1400 [11] R.J. Pelham Y.L. Wang Cell locomotion and focal adhesions are regulated by substrate flexibility (vol. 94, pp. 13661) Proceedings of the National Academy of Sciences of the United States of America 95 1997 13661 13665 [12] H.-B. Wang M. Dembo Y.-L. Wang Substrate flexibility regulates growth and apoptosis of normal but not transformed cells Am. J. Physiol. Cell Physiol. 279 2000 C1345 C1350 [13] C.M. Lo H.B. Wang M. Dembo Y.L. Wang Cell movement is guided by the rigidity of the substrate Biophys. J. 79 2000 144 152 [14] A.J. Engler M. Berry H.L. Sweeney D.E. Discher Substrate compliance alters human mesenchymal stem cell morphology Mol. Biol. Cell 15 2004 298A [15] Y. Ni M.Y.M. Chiang Cell morphology and migration linked to substrate rigidity Soft Matter 3 2007 1285 1292 [16] N. Yamamura R. Sudo M. Ikeda K. Tanishita Effects of the mechanical properties of collagen gel on the in vitro formation of microvessel networks by endothelial cells Tissue Eng. 13 2007 1443 1453 [17] C.S. Chen Mechanotransduction – a field pulling together? J. Cell Sci. 121 2008 3285 3292 [18] A.J. Engler M.A. Griffin S. Sen C.G. Bonnemann H.L. Sweeney D.E. Discher Myotubes differentiate optimally on substrates with tissue-like stiffness: pathological implications for soft or stiff microenvironments J. Cell Biol. 166 2004 877 887 [19] J.P. Winer P.A. Janmey M.E. McCormick M. Funaki Bone marrow-derived human mesenchymal stem cells become quiescent on soft substrates but remain responsive to chemical or mechanical stimuli Tissue Eng. Part A 15 2009 147 154 [20] S. Sen A. Engler D. Discher Matrix strains induced by cells: computing how far cells can feel Cell Mol. Biol. 2 2009 39 48 [21] J.M. Maloney E.B. Walton C.M. Bruce K.J. Van Vliet Influence of finite thickness and stiffness on cellular adhesion-induced deformation of compliant substrata Phys. Rev. E (Stat. Nonlinear Soft Matter Phys.) 78 2008 041915 041923 [22] NKn. Rudolf Merkel Claudia M. Cesa Bernd. Hoffmann Cell force microscopy on elastic layers of finite thickness Biophys. J. 93 2007 3314 3323 [23] B. Amnon How deeply cells feel: methods for thin gels J. Phys. Condens. Matter 22 2010 194116 [24] J.F. Leterrier J. Käs J. Hartwig R. Vegners P.A. Janmey Mechanical effects of neurofilament cross-bridges J. Biol. Chem. 271 1996 15687 15694 [25] C. Storm J.J. Pastore F.C. MacKintosh T.C. Lubensky P.A. Janmey Nonlinear elasticity in biological gels Nature 435 2005 191 194 [26] J.P. Winer S. Oake P.A. Janmey Non-linear elasticity of extracellular matrices enables contractile cells to communicate local position and orientation PLoS ONE 4 2009 e6382 [27] B.O. Palsson S.N. Bhatia Tissue Engineering 2003 Pearson Prentice Hall Upper Saddle River, NJ [28] A.L. Plant K. Bhadriraju T.A. Spurlin J.T. Elliott Cell response to matrix mechanics: focus on collagen Biochim. Biophys. Acta (BBA) – Mol. Cell. Res. 1793 2009 893 902 [29] D.G. Wallace J. Rosenblatt Collagen gel systems for sustained delivery and tissue engineering Adv. Drug Deliver. Rev. 55 2003 1631 1649 [30] L.L. Natalia A.C. Oksana V.V. Ekaterina A.O. Dmitry I.K. Elena V.M. Olga I.F. Daria V.E. Natalia A.D. Stephen Progression of HCC in mice is associated with a downregulation in the expression of hepatocyte nuclear factors Hepatology 39 2004 1038 1047 [31] L.B. Adam Engler Cynthia Newman Alina Hategan Maureen Griffin Dennis Discher Substrate compliance versus ligand density in cell on gel responses Biophys. J. 86 2004 617 628 [32] G.H.M. Yourek J.J. Mao Cytoskeletal changes of mesenchymal stem cells during differentiation ASAIO J. 53 2007 219 228 [33] S.R. Peyton C.M. Ghajar C.B. Khatiwala A.J. Putnam The emergence of ECM mechanics and cytoskeletal tension as important regulators of cell function Cell Biochem. Biophys. 47 2007 300 320 [34] D. Vader A. Kabla D. Weitz L. Mahadevan Strain-induced alignment in collagen gels PLoS ONE 4 2009 e5902 [35] F.J. Byfield R.K. Reen T.-P. Shentu I. Levitan K.J. Gooch Endothelial actin and cell stiffness is modulated by substrate stiffness in 2D and 3D J. Biomech. 42 2009 1114 1119 [36] S. Vanni B.C. Lagerholm C. Otey D.L. Taylor F. Lanni Internet-based image analysis quantifies contractile behavior of individual fibroblasts inside model tissue Biophys. J. 84 2003 2715 2727