ARTICLE IN PRESS
Biomaterials 26 (2005) 4985–4992 www.elsevier.com/locate/biomaterials
The ﬁbroblast response to tubes exhibiting internal nanotopography Catherine Cecilia Berrya,, Matthew J Dalbya, David McCloya, Stanley Affrossmanb a
Centre for Cell Engineering, University of Glasgow, IBLS, Joseph Black Building University, Glasgow G12 8QQ, UK Department of Pure and Applied Chemistry, Thomas Graham Building, University of Strathclyde, Glasgow G1 1XL, UK
Received 12 October 2004; accepted 17 January 2005
Abstract The use of three-dimensional scaffolds in cell and tissue engineering is widespread; however, the use of such scaffolds, which bear additional cellular cues such as nanotopography, is as yet in its infancy. This paper details the novel fabrication of nylon tubes bearing nanotopography via polymer demixing, and reports that the topography greatly inﬂuenced ﬁbroblast adhesion, spreading, morphology and cytoskeletal organisation. The use of such frameworks that convey both the correct mechanical support for tissue formation and stimulate cells through topographical cues may pave the way for future production of intelligent materials and scaffolds. r 2005 Elsevier Ltd. All rights reserved. Keywords: Actin; Cell adhesion; Cell morphology; Cell spreading; Nanotopography; Vinculin
1. Introduction The interaction of cells with the surface of laboratorydesigned materials is of great importance to progress in implant technology and tissue engineering, and is thus of great relevance in biomedical research . It is becoming clear that the role of biomaterials in this ﬁeld is extending from mere mechanical support to actually using intelligent material surfaces capable of providing chemical and physical cues to guide cell attachment, differentiation and thus aid the eventual assembly of cells to form functional tissue. One possibility that shows promise is that of adding deﬁned topography to a material [2,3]. Micro-topography has been shown to present powerful cues to cells, altering adhesion, movement, morphology, apoptosis, macrophage activation and gene expression [4–6]. Patterning methods developed by the electronics research sector such as photolithography, Corresponding author. Tel.: +44 0141 3303550; fax: +44 0141 3303730. E-mail address: [email protected]
0142-9612/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2005.01.046
electron beam lithography and laser holography are now allowing cell engineers access to well-deﬁned nanotopographies . Indeed, there is growing evidence from recent research that cells do react to nanoscale surface features [8–10]. Whilst it is possible to produce nanotopographies with 5 nm resolution using the aforementioned techniques, such methods are costly and time consuming, particularly if large areas of pattern are required. Recent work within our group has also concentrated on the production of rapid and cheap nanotopography by polymer demixing. In this technique, blends of polymers, for example, polystyrene (PS) and poly(4-bromostyrene) (PBrS), spontaneously undergo phase separation during spin casting onto silicon wafers . By controlling the polymer concentration and the proportions of the polymers, different topographies can be produced; these range from pits to islands or ribbons of varying height and depth. X-ray photoelectron spectroscopy (XPS) and static secondary ion mass spectrometry (SIMS) have been used to determine the surface composition of the blends [12,13], showing that PS segregates to the surface on annealing the ﬁlms. This
ARTICLE IN PRESS 4986
C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992
means that despite the topography being formed by polymer blends, the cells only interact with a single chemistry on annealed substrates. Another polymer blend that gives topography in cast ﬁlms is polystyrenepoly(n-butylmethacrylate (PS/PnBMA). In this system, the PnBMA segregates spontaneously to cover the surface at ambient temperature, producing a single chemistry without annealing . To date, there have been studies observing the inﬂuence of nanoisland topography with various heights created using polymer demixing on a variety of human cells in culture, in particular endothelial and ﬁbroblast cells, indicating a range of responses [15–18]. Interestingly, the islands of smaller heights (e.g. 14 nm) tended to increase adhesion/proliferation/cytoskeletal development/gene expression, while those of larger height (e.g. 95 nm) induced a stellate cell morphology with a poorly formed cytoskeleton [15,16]. It is noted, however, that the vast majority of research utilising topographical cues is two-dimensional (2D) on the macro-scale, and can thus be criticised for being a poor representation of most topographies presented to cells in the body. We are thus currently trying to capitalise on our wealth of information on topography fabrication and to transfer this knowledge to a 3D situation. Current research on the production of 3D scaffolds exhibiting topography tends to rely on prefabrication of the topography on a ﬂat surface, followed by subsequent shaping of the material, for example rolling into a tube. Earlier this year, our group presented the ﬁrst paper indicating successful nanopatterning inside a glass tube with an internal diameter (ID) of 1 mm. There are currently many shapes of scaffold under investigation; however, the successful production of a tube exhibiting internal topography would provide very useful insight for a wide range of cell and tissue engineers, encompassing, for example, vascular, bone and tendon/ligament research, and also beneﬁt stent design, providing support during and after surgical anastomosis. This paper also reported an initial observation of ﬁbroblast behaviour on culturing inside the nanopatterned tubes, resulting in the cells adopting a stellate morphology and forming large clusters . The method of internal patterning using polymer demixing is shown here to be applicable to and has more recently been adapted for use with, standard commercially available nylon tubing. Nylon tubing is organic and ﬂexible and transparent and thus is a step towards using a more physiological material. The present study focuses on the response of human ﬁbroblasts (as a cell type representative of those a material would contact in vivo) to islands formed from a PS/PnBMA blend and presented in two tube sizes with IDs of 0.5 and 1.5 mm, respectively. Qualitative and quantitative adhesion and morphological observations were made using light microscopy, scanning electron microscopy (SEM) and
ﬂuorescent observation of the cytoskeleton (F-actin, btubulin and vinculin).
2. Materials and methods 2.1. Nylon tube preparation Surgical, non-toxic, standard grade nylon tubing (Jencons), 0.5 or 1.5 mm ID, was used as obtained. The puriﬁcation of the polymers has been described previously . A blend of 20% PS and 80% PnBMA was dissolved in toluene at a concentration of 2% w/w. The polymer blend solution was introduced into the tube, length 50–100 cm, via a syringe and the solution was then blown through the tube by nitrogen at a pressure of 11.5 p.s.i. The process is similar to spin casting and leaves a thin layer of solution on the substrate surface that loses solvent rapidly to give a polymer ﬁlm. With a suitable blend of polymers, the components phase separates and the polymer surface exhibits topographical features. There is less control over the ﬁlm deposition conditions than with spin casting, so the topography is less uniform than on a ﬂat substrate. Accordingly, the samples used for the cell studies were selected from the middle of the tubing. A control sample was made by passing a toluene solution of the homopolymer, PnBMA, through the tubing. Cells seeded into the tubing will contact a relatively smooth PnBMA surface in the control tubing or a rough surface covered by PnBMA in the blendcoated samples. For ease of comparison the four samples will be referred to as indicated in Table 1. 2.2. Cell culture InﬁnityTM Telomerase Immortalised human ﬁbroblasts (h-TERT-BJ1, Clonetech Laboratories, Inc., USA) were seeded via injection (1 105 cells per ml medium) inside 3 cm sections of both 0.5 and 1.5 mm diameter tube. Cells were also seeded into sections of Table 1 Details of sample acronyms and average nanofeature heights for further reference Sample type
Average feature height
1.5 mm ID control 0.5 mm ID control 1.5 mm ID internal topography 0.5 mm ID internal topography
1.5 C 0.5 C 1.5 IT
NA NA 90 nm
ID denotes internal diameter.
ARTICLE IN PRESS C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992
blank controls (i.e. tubes with identical internal chemistry, minus topography). The medium used was 71% Dulbeccos Modiﬁed Eagles Medium, 17.5% Medium 199 (both Sigma, UK), 10% foetal calf serum, 1.6% 200 mM L-glutamine, and 0.9% sodium pyruvate (all Life Technologies, UK). The cells were incubated at 37 1C within a 5% CO2 atmosphere for 24 h. 2.3. Cell morphology and quantification of cell area After 24 h culture, the cells were washed in phosphate buffered saline (PBS) and then ﬁxed in 4% formaldehyde/PBS for 15 min at 37 1C. The cells were subsequently stained for 2 min in 0.5% Coomassie blue in a methanol/acetic acid aqueous solution, and washed with double distilled water to remove excess dye. Samples could then be observed by light microscopy, and digital images were captured using a Hamamatsu Argus 20 for image processing. For cell circularity and shape area quantiﬁcation, ImageJ was used. This used automated detection of cell outline and calculated the number of pixels covered by cells. It should be noted that standardised illumination conditions are used throughout. Student’s t-test (for two samples, assuming unequal variances) was used to compare statistical signiﬁcance of the patterned tube against its respective control. Results of po0.05 were considered signiﬁcant. 2.4. Immunofluorescence for cytoskeletal and focal contact observation After 24 h culture, the cells were ﬁxed in 4% formaldehyde/PBS, with 1% sucrose at 37 1C for 15 min. Once ﬁxed, the cells were washed in PBS. The individual sections of tubing were then dissected through the centre to permit easier access to the cells for further stain penetration. The cells were then incubated in a permeabilising buffer (10.3 g sucrose, 0.292 g NaCl, 0.06 g MgCl2, 0.476 g Hepes buffer, 0.5 ml Triton X, in 100 ml distilled water, pH7.2) at 4 1C for 5 min, with subsequent incubation in 1% BSA/PBS at 37 1C for 5 min. This was followed by the addition of anti-b-tubulin, or antivinculin primary antibody (1:100 in 1% BSA/PBS; tub 2.1 or Hvin1 monoclonal anti-human raised in mouse, (IgG1) Sigma, UK) for 1 h at 37 1C. Simultaneously, rhodamine phalloidin was added for the duration of this incubation (1:100 in 1% BSA/PBS, Molecular Probes, OR, USA). The samples were next washed in 0.5% Tween 20/PBS (5 min 3) prior to a secondary antibody (1:50 in 1% BSA/PBS monoclonal horse anti-mouse (IgG), Vector Laboratories, UK), which was added for 1 h at 37 1C. Finally, after further Tween 20/PBS washes, an FITC conjugated streptavidin third layer was added (1:50 in 1% BSA/PBS, Vector Laboratories, UK) for 30 min at 4 1C. Samples were mounted in Vectorshield
ﬂuorescent mountant and viewed by ﬂuorescence microscopy (Zeiss Axiovert 200M). 2.5. Scanning electron microscopy The cells were ﬁxed with 1.5% gluteraldehyde (Sigma, UK) buffered in 0.1 M sodium cacodylate (Agar, UK) (4 1C, 1 h) after 24 h culture. The cells were then postﬁxed in 1% osmium tetroxide for 1 h and 1% tannic acid (both Agar, UK) was used as a mordant. Samples were dehydrated through a series of alcohol concentrations (20%, 30%, 40%, 50%, 60%, 70%), stained in 0.5% uranyl acetate, followed by further dehydration (90%, 96%, 100% alcohol). The ﬁnal dehydration was in hexamethyl-disilazane (Sigma, UK), followed by airdrying. Once dry, the samples were coated with gold before examination with a Hitachi S800 ﬁeld emission SEM at an accelerating voltage of 10 keV.
3. Results 3.1. Atomic force microscopy of internal nanotopography Examination by AFM showed that the control tube’s internal surface was nanometrically ﬂat. AFM measurements for the 1.5 IT and 0.5 IT show a topography consisting of nanometric islands with typical mean heights of ca. 90 and 40 nm, respectively (Fig. 1a,b). The percentage substrate area covered by the islands was 18 and 23, respectively, consistent with the PS component concentration of 20%. The average island height and number density of islands depend on several factors including the casting solution concentration, cast liquid ﬁlm thickness and rate of solvent removal. These factors can be readily varied when spin-casting on to ﬂat substrates. However, the experimental restrictions imposed when casting inside a narrow tube limits the range of attainable topographies. 3.2. Cell morphology and area Cell morphology is illustrated by Coomassie blue staining of cells in Fig. 2. Note that most cell images may appear slightly out of focus in areas due to the nature of cell curvature in the tubes. The ﬁbroblasts cultured in the 1.5 C and 0.5 C tubes are well spread, whilst the cells cultured in the 1 IT and 0.5 IT tubes have adopted a highly stellate morphology, with a more rounded cell body. Quantiﬁcation of cell circularity at 24 h culture reﬂected the Coomassie blue images, with ﬁbroblasts cultured in the 1.5 IT and 0.5 IT tubes proving to be more circular than those cultured on the respective controls (Fig. 3).
ARTICLE IN PRESS 4988
C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992
Fig. 1. (a) PS/PnBMA blend in 1.5 mm ID tubing. Above: AFM image, below: height distribution. (b) PS/PnBMA blend in 0.5 mm ID tubing. Above: AFM image, below: pixel height distribution.
Fig. 3. Graph showing the average cell circularity (perfect circle ¼ 1.0) for ﬁbroblasts cultured in the control and internal topography tubes (results are mean7SD, ¼ t-test, po0.01). Fig. 2. Coomassie blue light micrograph images of cells cultured for 24 h in 1.5 C tubes (a), 0.5 C tubes (b), 1.5 IT tubes (c) and 0.5 IT tubes (d). Note the difference in ﬁbroblast morphology between the control tubes and those exhibiting topography.
tubes, suggesting that the latter had not adhered and spread as well. 3.3. Cell cytoskeleton and focal contacts
Quantiﬁcation of cell area in Fig. 4 showed that the ﬁbroblasts cultured in control tubes had signiﬁcantly greater areas than those cultured in the 1.5 IT and 0.5 IT
Fibroblasts cultured in the control tubes indicated Factin stress ﬁbres forming in both the cytoplasm and the
ARTICLE IN PRESS C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992
Fig. 4. Graph showing the average cell area for ﬁbroblasts cultured in the control and internal topography tubes (results are mean7SD, ¼ t-test, po0.01).
F-actin mainly at the cell periphery (Fig. 5b1,d1). In both cases, the b-tubulin, although appearing less organised compared to controls, did exhibit radiating ﬁbres throughout the cell body, forming a structured cytoskeleton (Fig. 5b2,d2). With regard to focal contact formation, cells cultured in the control tubes formed clear dash-shaped contacts mainly around the cell periphery, with an organised actin cytoskeleton (Fig. 6a,c). Cells cultured in the 1.5 IT tubes also indicated reasonably formed contact at the cell periphery, with little actin organisation (Fig. 6b), while cells cultured in the 0.5 IT tubes exhibited very small and indistinct contacts, again with little actin organisation (Fig. 6d).
3.4. Cell interaction with internal nanotopography via SEM SEM observation of cell morphology in the 1.5 C and 0.5 C tubes indicated ﬂattened cell morphology with
Fig. 5. F-actin and b-tubulin staining for ﬁbroblasts cultured for 24 h in 1.5 C tubes (a1, a2), 1.5 IT tubes (b1, b2), 0.5 C tubes (c1, c2) and 0.5 IT tubes (d1, d2), respectively.
cell periphery. In addition, the b-tubulin appeared well organised, forming radiating ﬁbres out from the organising centre beside the nucleus (Fig. 5a1,a2 c1,c2). Fibroblasts cultured in the 1.5 IT and 0.5 IT tubes, however, showed poorly organised and punctate
Fig. 6. F-actin and vinculin staining for ﬁbroblasts cultured for 24 h in 1.5 C tubes (a1, a2), 1.5 IT tubes (b1, b2), 0.5 C tubes (c1, c2) and 0.5 IT tubes (d1, d2), respectively.
ARTICLE IN PRESS 4990
C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992
Fig. 7. SEM micrographs of ﬁbroblasts cultured for 24 h in 1.5 C and 0.5 C tubes (a, b), 1.5 IT tubes (c, e) and 0.5 IT tubes (d, f), respectively.
smooth cell edges (Fig. 7a,b, respectively). This was clearly different for the cell morphology exhibited by ﬁbroblasts cultured in the 1.5 IT and 0.5 IT tubes, where the cell membrane appeared rufﬂed, with numerous ﬁlopodia interacting with the island topography (denoted by arrowheads in Fig. 7c,d). These ﬁlopodia were exceptionally long compared to the cell body dimensions, as highlighted in Fig. 7e and f, and usually terminated on one of the nanometric islands.
4. Discussion Our group very recently reported on the ﬁrst successful introduction of nanotopography into a 3D tube . This opens up a wide range of possible biomedical applications in the cell and tissue engineering arena, such as vascular, bone and tendon/ligament research. We know, from previous work in our group, that we can use micro- and nanotopography to induce very speciﬁc cell responses ranging from preventing adhesion  to increasing adhesion/proliferation and gene expression of a range of proteins important in cell survival and growth [4,21]. This paper therefore details the ﬁrst logical stage in studying potential applications
of tubes with internal topography by gauging cell interaction with the topography. The cell morphology images and quantiﬁcation data indicate a very obvious difference in cell size and shape when cultured in control tubes compared to those with topography. The latter reveal a smaller cell area and are more rounded/circular with stellate cell morphology, suggesting poor cell spreading. Cell spreading and resultant morphology is important for many cells as it is necessary for cell division, and ﬁbroblastic cells that cannot attach and spread properly have been linked to a particular type of apoptosis termed anoikis, which originates from poor interactions of cells with their external environment [22,23]. It has been previously reported that ﬁbroblast spreading and morphology can be controlled using polymer demixing on coverslips, depending on the island height produced (, for a comprehensive review see ). Brieﬂy, 13 nm islands increased cell spreading, with the opposite occurring in response to the 95 nm islands, when compared to ﬂat controls. A similar trend was noted for cell proliferation, believed to be directly linked to the poor cell spreading. SEM visualisation of the cell/island interaction indicated that the larger the islands the more the ﬁlopodial interaction, eventually
ARTICLE IN PRESS C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992
resulting in stellate-shaped cells on the 95 nm islands. The two tube sizes used in this study exhibit differentsized topographies, namely heights of 40 and 90 nm for the 0.5 IT and 1.5 IT, respectively. The results presented here show that the cells are reacting to both the topographies in a similar fashion to the above 95 nm islands on coverslips, i.e. with reduced cell spreading and rounded cell morphologies. With regard to cell cytoskeleton, cells in control tubes had clearly deﬁned actin ﬁbres, with prominent stress ﬁbres, and tubulin ﬁbres radiating out for the centre of the cell, as opposed to the cells exposed to the topographies which exhibited poor actin organisation, but reasonably structured tubulin. Changes in cell shape, anchorage and motility are all associated with the dynamic reorganisation of the ﬁlament arrays that make up the actin cytoskeleton. Cells cultured in the IT tubes exhibit punctate actin throughout the cell body, with large amounts around the cell periphery. Similar types of actin organisation are reportedly found in cells that have relaxed their cytoskeleton . Activation of individual members of the Rho family of small G proteins is known to modulate the organisation of actin ﬁlaments in cells, with Rho activation resulting in the formation of stress ﬁbres and focal contacts, Rac inducing lamellipodia and Cdc42 inducing ﬁlopodia formation [27,28]. It could therefore be proposed that the topographically induced alteration in morphology/cytoskeleton in response to the internal tube nano-islands is due to changes in these proteins. Indeed, a microarray study of cell reaction to 13 nm high islands on ﬂat glass coverslips, previously reported to increase cell adhesion, spreading and proliferation, showed upregulations for Ras, Rab and Rho G proteins . As mentioned above, the microtubule network in cells cultured in the IT tubes remains relatively well structured, with apparent ﬁbres. The most prominent roles for microtubules in the cell are the transport of vesicles and organelles in and out of the cell and also in cell division (forming the spindle) . Thus, the fact that tubulin is relatively organised may explain the fact that although the cells are compromised with regard to morphology and size, they are viable and metabolising (shown in viability staining, data not included here). Fibroblast spreading under normal conditions entails a cycle of cell–substrate contact formation, whereby new focal contacts form beneath lamellipodium and develop into focal complexes/adhesions as the cell body advances, thereby allowing the cell to spread . Fibroblasts in the control tubes exhibit dash-shaped adhesions throughout the cell, reﬂecting their wellspread morphology, whereas ﬁbroblasts in the 0.5 IT tubes have very small adhesions reﬂecting their smallersized morphology. These adhesion complexes are known to vary in size and organization; thus, the formation and
remodelling of focal contacts/adhesions is a dynamic process under the regulation of protein tyrosine kinases and members of the Rho family. Brieﬂy, Rac functions to signal the creation of new substrate contacts at the cell front (associated with the rufﬂing of lamellipodia), whereas Rho serves in the maturation of existing contacts . Thus, if the Rho family of proteins were involved in the changes observed in cell size and morphology, then it follows that they may also be inﬂuencing formation of adhesions.
5. Conclusion In conclusion, this paper presents a simple method of preparing tubes with an internal nanotopography that is capable of inducing large changes in ﬁbroblast morphology, size and adhesion, with clear interaction between the cell ﬁlopodia and the presented nanofeatures. Interestingly, there were no differences observed when comparing the two different diameters of tubing; both appeared to induce very similar cell responses. Although several proposals as to the changes observed are given, cell response needs to be observed at the molecular level to further elucidate the mechanisms of cell reaction. Whilst at an early developmental stage, we envisage that control of cells within tubes will present potential for tissue engineering scaffolds such as with stents and conduits for vascular and nerve regeneration.
References  Grifﬁth LG, Naughton G. Tissue engineering—current challenges and expanding opportunities. Science 2002;295:1009–14.  Curtis ASG, Wilkinson CDW. Topographical control of cells. Biomaterials 1997;18:1573.  Curtis ASG, Wilkinson CDW. Reactions of cells to topography. J Biomed Sci Polym Ed 1998;9:1313.  Berry CC, Campbell G, Spadiccino A, Robertson M, Curtis AS. The inﬂuence of microscale topography on ﬁbroblast attachment and motility. Biomaterials 2004;25:5781–8.  Clark P, Connolly P, Curtis ASG, Dow JAT, Wilkinson CDW. Topographical control of cell behaviour: simple step cues. Development 1987;99:439.  Britland S, Morgan H, Wojciack-Stothard B, Riehle M, Curtis ASG, Wilkinson CDW. Synergistic and hierarchical adhesive and topographical guidance of BHK cells. Exp Cell Res 1996;228:313–25.  Wilkinson CDW, Riehle M, Wood MA, Gallacher J, Curtis ASG. The use of materials patterned on a nano- and micro-metric scale in cellular engineering. Mater Sci Eng 2002;19:263–9.  Desai TA. Micro- and nanoscale structures for tissue engineering constructs. Med Eng Phys 2000;22:595–606.  Dalby MJ, Riehle MO, Sutherland DS, Agheli H, Curtis ASG. Changes in ﬁbroblast morphology in response to nano-columns produced by colloidal lithography. Biomaterials 2004;25:5415–22.  Dalby MJ, Riehle MO, Johnstone H, Affrossman S, Curtis AS. Investigating the limits of ﬁlopodial sensing: a brief report using
ARTICLE IN PRESS 4992
C.C. Berry et al. / Biomaterials 26 (2005) 4985–4992 SEM to image the interaction between 10 nm high nanotopography and ﬁbroblast ﬁlopodia. Cell Biol Int 2004;28:229–36. Affrossman S, Henn G, O’Neil SA, Pethrick RA, Stamm M. Surface topography and composition of deuterated polystyrenepoly(bromostyrene) blends. Macromolecules 1996;29:5010–6. Affrossman S, O’Neil SA, Stamm M. Topography and surface composition of thin ﬁlms of blends of polystyrene with brominated polystyrene: effects of varying the degree of bromination and annealing. Macromolecules 1998;31:6280–8. Affrossman S, Stamm M. The effect of molecular weight on the topography of thin ﬁlms of blends of poly(4-bromostyrene) and polystyrene. Colloid Polym Sci 2000;278:888–93. Affrossman S, Jerome R, O’Neill SA, Schmitt T, Stamm M. Surface structure of thin ﬁlm blends of polystyrene and poly(nbutyl methacrylate). Colloid Polym Sci 2000;278:993–9. Dalby MJ, Riehle M, Johnstone HJH, Affrossman S, Curtis ASG. Polymer demixed nano-topography: control of ﬁbroblast spreading and proliferation. Tiss Eng 2002;8:1099–108. Dalby MJ, Yarwood SJ, Riehle M, Johnstone HJH, Affrossman S, Curtis ASG. Increasing ﬁbroblast response to materials using nano-topography: morphological and genetic measurements of cell response to 13 nm high polymer demixed islands. Exp Cell Res 2002;276:1–9. Dalby MJ, Childs S, Riehle M, Johnstone HJH, Affrossman S. Fibroblast reaction to island topography: changes in cytoskeleton and morphology with time. Biomaterials 2003;24:927–35. Dalby MJ, Giannaras D, Riehle M, Gadegaard N, Affrossman S, Curtis ASG. Rapid ﬁbroblast adhesion to 27 nm high polymer demixed nano-topography. Biomaterials 2004;25:77–83. Gadegaard N, Dalby MJ, Riehle MO, Curtis ASG, Affrossman S. Tubes with controllable internal nanotopography. Adv Mater 2004;16(20):1857–9.
 Gallagher JO, McGee KF, Wilkinson CDW, Riehle MO. Interaction of animal cells with ordered topography. IEEE Trans Nanobiosci 2002;1:24–8.  Dalby MJ, Riehle MO, Yarwood SJ, Wilkinson CD, Curtis ASG. Nucleus alignment and cell signaling in ﬁbroblasts: response to a micro-grooved topography. Exp Cell Res 2003;284:274–82.  Meredith JE, Fazeli B, Schwartz MA. The extracellular matrix as a cell survival factor. Mol Biol Cell 1993;4:953–61.  Firsch SM, Francis H. Disruption of endothelial cell–matrix interactions induces apoptosis. J Cell Biol 1994;124:619–26.  Dalby MJ, Rielhe MO, Johnstone HJH, Affrossman S, Curtis ASG. Nonadhesive nanotopography: ﬁbroblast response to poly(n-butyl methacrylate)-poly(styrene) demixed surface features. J Biomed Mater Res 2003;67A:1025–31.  Dalby MJ, Pasqui D, Affrossman S. Cell response to nano-islands produced by polymer demixing: a brief review. IEEE Proc Nanobiotech 2004;151:53–61.  Small JV, Rottner K, Kaverina I. Functional design in the actin cytoskeleton. Curr Opin Cell Biol 1999;11:54–60.  Amano M, Chihara K, Kimura K, Fukata Y, Nakamura N, Matsuura Y, Kaibuchi K. Formation of actin stress ﬁbres and focal adhesions enhanced by Rho-kinase. Science 1997;275: 1308–11.  Aspenstrom P. The Rho GTPases have multiple effects on the actin cytoskeleton. Exp Cell Res 1999;246:20–5.  Elbaum M, Chausovsky A, Levy ET, Shtutman M, Bershadsky AD. Microtubule involvement in regulating cell contractility and adhesion-dependent signalling: a possible mechanism for polarisation of cell motility. Biochem Soc Symp 1999;65:147–72.  Rottner K, Hall A, Small JV. Interplay between Rac and Rho in the control of substrate contact dynamics. Curr Biol 1999;9:640–8.