Microporous “Honeycomb” Films Support Enhanced Bone Formation In Vitro

Abstract

Substrate topography influences cell adhesion, proliferation, and differentiation. In this study, poly (ɛ-caprolactone) (PCL) films with a well-defined honeycomb structure of porosity 3–4, 5–6, 10–11, or 15–16 μm were contrasted with flat surfaces for their ability to support primary rat osteoblast adhesion and mineralized extracellular matrix deposition in vitro. Immunofluorescent visualization of vinculin and rhodamine phalloidin binding of actin were used to investigate cell adhesion and morphology. Localization of the alkaline phosphatase activity and Alizarin Red staining were performed to assess the osteoblast activity and deposition of a mineralized matrix. Scanning electron microscopy together with energy-dispersive X-ray spectroscopy was used to provide morphological analysis of cell–film interactions, the deposited matrix, and elemental analysis of the mineralized structures. After 24 h of culture, there were no differences in cell numbers on porous or flat PCL surfaces, but there were changes in cell morphology. Osteoblasts on honeycomb films had a smaller surface area and were less circular than cells on flat PCL. Analysis of cells cultured for 35 days under osteogenic conditions revealed that osteoblasts on all substrates acquired alkaline phosphatase activity, but levels of mineralized matrix were increased on films with 3–4-μm pore sizes. The bone-like matrix with a Ca:P ratio of 1.69±0.08 could be identified in larger areas often aligning with substrate topography. In addition, smaller spherical deposits (0.5–1 μm in diameter) with a Ca:P ratio of 1.3±0.08 were observed at the surface and particularly within the pores of the PCL film. Localization of vinculin showed significant decreases in the number of focal adhesion structures per unit cell area on 5–6, 10–11, and 15–16-μm surfaces compared to flat PCL, while focal complexes with a smaller area (0–2 μm²) were more abundant on 3–4 and 5–6-μm surfaces. Observation of cell interaction with these surfaces identified cytoplasmic protrusions that extended into and sealed the pores of these PCL films creating an extracellular space in which, the conditions could influence the deposition and formation of the mineralized matrix.

Introduction

It is well established that cell growth and activity can be influenced by interactions with surface topography and structure. Micro/nanofabrication approaches have been used to produce well-defined and characterized surfaces to act as cell culture substrates and assess their role in regulating cell adhesion, morphology, and gene expression.^1–5 These studies have revealed that the control of cytoskeletal organization and, in particular, the formation of adhesion structures and stress fibers can be linked to topographical features impacting on cell guidance and migration.^6,7 Other studies have shown that surfaces can be designed to influence signal transduction pathways and gene expression^8–11 leading to the hypothesis that altered internal cell tensioning and nuclear morphology may play a part in surface-induced regulation of cell activity.^12,13

Biomaterial scientists have long understood that the combination of surface chemistry and structure have the potential to be tailored to ensure that an appropriate tissue response is elicited by an implant. This is particularly evident in orthopedics where surface structuring has been used to improve osseointegration, the process of bone growth directly onto the biomaterial, ensuring long-term fixation and functional effectiveness.^14–17 Deposition of bone is achieved by osteoblasts¹⁸ and these cells deposit a proteinaceous matrix that is predominantly type I collagen, and then mediate mineralization by the release of matrix vesicles¹⁹ that drive the formation of hydroxyapatite crystal formation within the protein fibrils.

Previous studies have demonstrated that osteoblast adhesion, proliferation, and differentiation are enhanced by textured surfaces.^20,21 However, increased surface roughness may also negatively influence cell adhesion and bone formation.^22,23 The gene expression profile, particularly those involved in mesenchymal cell differentiation, has been shown in bone marrow stromal cells to be regulated by surface topography.²⁴ In addition, microtopography also modulates how osteoblast cells respond to osteotropic regulatory factors such as 17β-estradiol,²⁵ 1, 25(OH)₂D₃,⁸ and BMP-2.²⁶ It is therefore clear that biomaterial surfaces have the potential to influence cell adhesion, profile of gene expression, and response to growth factors and hormones.

What remains unclear are the exact architectural parameters that influence bone cell function and the mechanisms that control the cell phenotype. Detailed studies on the influence of regular porous films will help clarify the mechanisms underlying the bone cell activity on biomaterials. We have previously found that regular structures are formed during the casting of biodegradable polymer solutions on solid surfaces; for example, self-organized honeycomb-patterned (honeycomb) films^27–33 with highly regular porous structures can be prepared under humid casting conditions. In this study, planar substrates with highly controlled honeycomb surface features were created in poly (ɛ-caprolactone) (PCL) and assessed for their ability to support osteoblast adhesion, spreading, cytoskeleton and focal adhesion complex formation, and long-term growth under osteogenic culture conditions.

Materials and Methods

Materials

Films were constructed using PCL (MW=70,000–100,000) and a copolymer of didecylacrylamide and ω-carboxyhexylacrylamide (CAP; MW=22,000). The synthesis of CAP has previously been described³⁴ and was used as an emulsifying agent in our studies. Polymer surfaces for cell culture were prepared on cover glasses (diameter: 15 mm; Matsunami Glass Industrial Ltd.). Corning^® Costar^® (Corning, Inc.) cell culture plasticware was used throughout these studies.

Preparation of flat and porous films

The porous honeycomb films of PCL were fabricated as previously described^27–32 using water droplets as templates for regular pores. In brief, PCL and CAP (10:1 wt%) were dissolved in chloroform at a concentration of 5 g/L and the polymer mix layered over the cover glasses in a humid environment. The pore sizes of the films were controlled using the following casting conditions: a temperature of 22°C, 85% humidity, a gas flow rate of 1.0 L/min, and a casting volume of 1–15 mL. Films typically exhibited 50% porosity and the pore size was principally controlled by the casting volume with 1 mL giving a film with 3–4-μm pores, 5 mL a film with 5–6-μm pores, 10 mL a film with 10–11-μm pores, and 15 mL a film with 15–16-μm pores. Film thicknesses were equivalent to the scale of the porosity, for example, the film with 3–4-μm pores was ∼3-μm thick, while the 10–11-μm film had a thickness of 10 μm. Flat polymer layers of PCL/CAP were created by casting the solution over the cover glasses and then washing in 1-propanol for 12 min to remove CAP.

Cell culture

Primary calvarial rat osteoblasts isolated as previously described³⁵ were cultured in the Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% (v/v) fetal calf serum (Sigma), and 2 mMol glutamine (Sigma). Cells from 2nd or 3rd passage were plated onto plane glass, flat PCL, or the porous PCL films at 10,000 cells/well (5000 cells/cm²) in short-term experiments (up to 72 h) to investigate cell morphology and at 20,000 cells/well (10,000 cells/cm²) in studies to osteoblast matrix deposition and mineralization. In long-term experiments, the culture media were exchanged after 48 h to include 5 mM β-glycerophosphate (Sigma), 100 μg/mL L-ascorbic acid phosphate (Wako Ltd.), and 10 nM dexamethasone (Sigma) to induce osteoblastic differentiation over periods up to 35 days. Throughout the course of the experiments, the cultures were fed every 3 days with osteogenic media. All cell cultures were performed at 37°C in 5% CO₂, humidified atmosphere.

Immuno/histochemical analysis

To visualize vinculin, cells were fixed in 4% (w/v) parformaldehyde, washed, and incubated with a primary antibody (V9131; Sigma) at 1:200 for 2 h at room temperature. After a wash step, a FITC-goat anti-mouse antibody (Zymed Laboratories) in PBS containing 50 μg/mL TRITC-phalloidin (Sigma) was added to the samples for 1 h in a dark chamber. Following a further wash step, cells were overlayered with VectaShield containing DAPI (Vector Labs) and mounted for microscopy. The alkaline phosphatase activity was localized with Naphthol AS MX-PO4 and Fast Blue RR salt and the mineralized matrix with Alizarin Red staining. Digital images were captured on a Leica DMLB microscope and analyzed with ImageJ. To study cell morphology, individual cells were manually outlined using a digital graphics tablet and the area calculated together with circularity, which was defined as 4π (area/perimeter²). To assess the mineralized matrix, red channel images were split from the randomly acquired RGB images and thresholded. The area of the mineralized matrix was then determined for each binary image. The assessment of vinculin-positive structures was performed by thresholding the grayscale image of stained cells, manual removal of perinuclear structures, then creation of a ROI equivalent to an individual cell as outlined above. Vinculin-positive staining within the ROI was then quantified using the menu command Analyze—> Analyze particles. In addition, confocal laser scanning microscopy (CSLM; Olympus FV-300) was also used to observe the specimens.

Scanning electron microscopy

Samples that had been in culture under osteogenic conditions for 35 days were fixed in 1.25% gluteraldehyde (Polysciences Europe GmbH) in PBS for 30 min. Specimens were then washed in PBS and water before dehydration in graded alcohol through to 100% ethanol and then t-butylalcohol. Critical point drying was then carried out and the samples coated with carbon. All analyses were performed on a scanning electron microscope (Hitachi S-4800) with Energy-dispersive X-ray spectroscopy (EDX) capability. EDX analysis was performed on eight independent sites of matrix mineralization at the surface of the films and eight points inside the pores of the films. Ca/P ratios were calculated and a mean±standard deviation is presented.

Statistical analysis

Similar findings were observed in at least three independent experiments for each experimental approach with data from a single experiment being presented as representative of these studies. Statistical significance between data sets was calculated using either an ANOVA with a Tukey post-test or Kruskal–Wallis one-way analysis of variance with a Bonferroni post-test. Differences were considered significant when p<0.05.

Results

Porous honeycomb films control osteoblast morphology

Primary rat osteoblasts were seeded onto the glass and PCL substrates and cultured for between 16–18 h. Samples were then investigated using immunofluorescence microscopy to localize vinculin, cytoskeletal actin, and nuclear DNA. Cells were observed to adhere, form focal adhesion complexes, and robust actin cytoskeletal structures on all of the substrates investigated in this study (Fig. 1). Analysis of rat osteoblasts cultured on glass, flat and porous films showed that a similar number of cells could be observed after 16–18 h (Fig. 2a). There were, however, significant differences in cell morphology between the surfaces (Fig. 2b, c). Cells cultured on flat PCL were spread more extensively and rounder than cells plated on honeycomb films. There were no significant differences in terms of the cell area or roundness between the different porous surfaces.

FIG. 1.

Primary rat osteoblasts adhere and spread on porous PCL films. Cells (5000 cells/cm²) were seeded on (a) glass coverslips; (b) flat PCL films or PCL films with pore dimensions (c) 3–4 μm; (d) 4–5 μm; (e) 10–11 μm; or (f) 15–16 μm and cultured for 24 h. Cells were then fixed and vinculin (green) localized to cell–matrix contacts by immunofluorescence and actin cytoskeleton (red) visualized with TRITC-phalloidin. Nuclei (blue) were identified by DAPI staining. Scale bar=50 μm. PCL, poly (ɛ-caprolactone). Color images available online at www.liebertpub.com/tea

FIG. 2.

Analysis of immunofluorescence images of cells at 24 h on the PCL surfaces assessed for (a) cell number, (b) cell area, and (c) cell circularity. Fifty cells were investigated for both cell area and circularity, while the cell number was assessed in 10 random fields. Data from analysis of an individual experiment are presented as a bar (25th and 75th percentile) and the whisker (10th and 90th percentile) plot showing the median (single black line) and mean (thick red line). Similar results were obtained in three independent experiments. ANOVA with the Tukey post-test was used to determine statistical significance *p<0.05 compared to all other honeycomb films. Color images available online at www.liebertpub.com/tea

Porous PCL films support enhanced mineralized matrix deposition by osteoblasts

Rat osteoblasts were cultured on the substrates for 35 days under osteogenic conditions. Samples were then processed to localize the alkaline phosphatase activity and the formation of the mineralized matrix (Figs. 3a–e and 4a). On all of the different substrates investigated, there was evidence of an extensive alkaline phosphatase activity with cell aggregates displaying intense alkaline phosphatase activity typical of cells undergoing osteogenesis. Alizarin Red S staining revealed the deposition of the mineralized matrix and identified increased abundance of bone-like nodules on the porous surfaces compared to flat PCL. Image analysis of Alizarin Red S staining identified the significant levels of calcium deposition on the 3–4-μm porous PCL film compared to flat PCL with a progressive decrease in the levels of the mineralized matrix as the pore size increased (Fig. 3f).

FIG. 3.

PCL films support differential mineralized matrix deposition by osteoblasts in vitro. Primary rat ostoblasts (20,000 cells per well/10,000 cells per cm²) were seeded onto (a) flat PCL films or PCL films with pore dimensions (b) 3–4 μm; (c) 4–5 μm; (d) 10–11 μm; or (e) 15–16 μm and cultured under ostogenic conditions for 35 days. Samples were then fixed, Alizarin Red S stained to identify the mineralized matrix and alkaline phosphatase activity localized with a Brilliant Blue substrate. Scale bar=250 μm. The bone nodule area was quantitated by image analysis (f). Data are presented from a single experiment that is representative of three independent experiments, showing the mean of nine samples with statistical analysis by ANOVA with a Tukey post-test, *p<0.05 compared to flat and 15–16-μm films. Color images available online at www.liebertpub.com/tea

FIG. 4.

Deposition of a mineralized matrix can be influenced by film topography. After 35 days in osteogenic culture on a honeycomb film with porosity 4–5 μm, (a) Alizarin Red stained calcified matrix could be observed following the macrostructure of the porous film (black arrows). Scale bar=40 μm. (b) SEM analysis demonstrated the accumulation of fibrous matrix scaffolds on the surface and particularly at the rim of the PCL films (white arrow). In addition, spherical aggregates of a matrix were observed on the surface and within the pores of the films (gray arrow). Scale bar=40 μm (c) EDX analysis [single area denoted by red cross hair on (b)] gave a Ca/P ratio of 1.65. EDX, energy-dispersive X-ray spectroscopy; SEM, scanning electron microscope. Color images available online at www.liebertpub.com/tea

On flat PCL surfaces, Alizarin Red S staining was restricted to globular deposits with little obvious higher order structure other than the aggregation of surrounding cells. On the porous PCL films, it was observed that surface topography could guide matrix deposition with the mineralized matrix mapping to the rims of pores in some areas (Fig. 4a, b). Furthermore, EDX analysis (Fig. 4c) of these deposits typically gave a Ca/P ratio of 1.69±0.08, n=8 (independent samples).

Also evident on the surface of the PCL films and particularly within the pores were spherical matrix deposits, typically 0.5–1 μm in diameter (Figs. 4b and 5a, b). EDX analysis again identified that these structures consisted predominantly of calcium and phosphate (Fig. 5c, d) with a Ca/P ratio of 1.3±0.08, n=8 (independent samples).

FIG. 5.

Abundant accumulation of calcospherites within the pores of films with small dimensions. SEM analysis of 4–5-μm films (a) and (b) showed evidence of the mineralized matrix on the surface (white arrow) and within the pores (gray arrow). Scale bar=10 μm. EDX analysis revealed (c) the Ca/P ratio of 1.69 for the surface matrix (red cross hairs) and (d) a Ca/P ratio of 1.3 for a typical calcospherite (yellow cross hairs). Color images available online at www.liebertpub.com/tea

Cell attachment mechanisms on PCL films

It has previously been reported that cell adhesion, as reflected by the size and abundance of vinculin-containing focal adhesion complexes, influences cell differentiation and bone formation.^9,24,36 Experiments were therefore performed that looked at vinculin-positive focal adhesion complexes in cells cultured on flat PCL films and also the porous substrates (Fig. 6a, b). Image analysis was performed assessing the abundance of focal structures (1–5 μm² in size) in cells on each of the surfaces (Fig. 6c). These data show that when normalized for cell area, the number of focal adhesions was significantly reduced in cells grown on the surfaces with 5–6, 10–11, and 15–16-μm pores compared to the flat and 3–4-μm substrates. Further analysis identified the size distribution of focal complexes in cells on each of the surfaces (Fig. 7). On 3–4 and 5–6-μm pore size PCL, there was a clear shift to increased abundance of focal structures in the order of 0–2 μm², whereas cells on 10–11 and 15–16-μm substrates had elevated levels in the size range 2–4 μm². The vinculin-positive structures that were assessed in terms of focal complexes were at the surface of the PCL films, but our studies with rat osteoblasts also identified that immunoreactivity could also be found localized to the structure of the pore (Fig. 8a) extending the cell membrane inside the porous film (Fig. 8b).

FIG. 6.

Focal complexes exhibit distinct patterns of distribution on porous PCL. Primary rat osteoblasts adhere and spread on porous PCL films. Cells (11,000 cells/cm²) were seeded on (a) flat PCL films or PCL films with pore dimensions 3–4 μm (b); 4–5 μm; 10–11 or 15–16 μm and cultured for 24 h. Cells were then fixed and vinculin (green) localized to cell–matrix contacts by immunofluorescence (e.g., white arrows) and actin cytoskeleton (red) visualized with TRITC-phalloidin. Nuclei (blue) were identified by DAPI staining. Scale bar=40 μm. (c) For each experimental group, image analysis identified the numbers of focal complexes (1–8 μm²) within an individual cell and normalized this for cell area. Data from a single experiment are presented and are representative of three independent experiments. Values are the mean of eight samples from a single experiment with statistical analysis by ANOVA with a Tukey post-test, *p<0.05 compared to flat and 3–4-μm films. Color images available online at www.liebertpub.com/tea

FIG. 7.

Focal complex size distributions are influenced by the PCL film structure. Image analysis identified the areas of focal complexes formed by 20 individual cells cultured on flat PCL and each of the porous PCL substrates. The numbers of vinculin-positive structures falling within the size boundaries 0–2, 2–4, 4–8, 8–15, and 15–30 μm² were determined and plotted for (a) glass and (b–f) all films. ** Indicates significant differences in the abundance of focal complexes in cells on a given surface in the range 2–4 μm² compared to other size groups, using Kruskal–Wallis one-way analysis of variance with a Bonferroni post-test. Color images available online at www.liebertpub.com/tea

FIG. 8.

Cell membrane protrusions were identified extending into the porous PCL structure. Staining of cells with anti-vinculin (green) localized to cell–matrix contacts by immunofluorescence and actin cytoskeleton (red) was visualized with TRITC-phalloidin. Nuclei (blue) were identified by DAPI staining. (a) Epifluorescence microscopy identified increased concentrations of immunoreactivity (green) beneath cells that mapped to and were defined by the pores of the PCL substrate (e.g., white arrows) Scale bar=20 μm. (b) CLSM with cross-section image analysis confirmed that that vinculin positivity was in cytoplasmic projection into the pores of the PCL films (white arrows). Color images available online at www.liebertpub.com/tea

Discussion

Previous studies have demonstrated how topography at the nano- and microscale has the potential to influence osteoblast differentiation and bone formation.^5,9,17,24,37 What is less well understood are the molecular mechanisms that underlie this observation. Much of the focus has been on the role that cell attachment mechanisms may play in the subsequent control of cell activity, correlating topographically influenced formation of cell adhesion structures such as focal adhesion complexes with changes in the cell structure and function.^12,13 Evidence suggests that abundant large vinculin-positive adhesion structures and robust cytoskeleton can be linked with cell differentiation³⁶ and this may well be mechanistically analogous to the cellular response to substrate stiffness.³⁸ These observations have led to the suggestion that cytoskeletal tensioning in continuum with nuclear scaffolding proteins may directly influence the transcriptional activity and thus alter the cell activity.^39,40

The studies presented here used topographically distinct surfaces with defined and highly reproducible structures that ranged in size from subcellular (3–4 μm) to cellular (15–16 μm) to investigate their ability to influence osteoblast interactions with the surface and the formation of the mineralized matrix. There was little difference in the number of cells that adhered to these surfaces at 24 h of culture (Fig. 2a), but parameters of cell morphology (area and circularity) illustrated that cells on the porous films were significantly smaller and less circular than those on the flat film (Fig. 2b, c). Analysis of individual cells confirmed that the surface topographical features of the films used in this study were influencing cytoplasmic projections and spreading of the cells (Fig. 1).

Osteoblasts were cultured on the flat and porous films for periods of up to 35 days under osteogenic culture conditions to induce osteoblast differentiation. On all surfaces investigated, there were abundant numbers of cells that were positive for alkaline phosphatase and evidence of mineralized matrix deposits. In particular, an assessment of the area of calcified bone-like nodules revealed that they were increased on the 3–4 and 5–6-μm porous films compared to the other surfaces investigated in this study (Fig. 3). Observations of the mineralized matrix structures revealed that morphologically they could be divided into two classes: one, where the gross structure had, in part, been dictated by the surface architecture of the substrate material (Fig. 4) or a second spherical morphology that appeared not to be physically related to surface topography (Fig. 5).

The larger deposits of the mineralized matrix exhibited a Ca:P ratio of 1.69, which is close to the nominal value of pure hydroxapatite (1.67) and typical of the values observed for hydroxyapatite formed in vitro by osteoblasts. It was also observed that the macrostructural organization of these bone-like nodules often conformed to the topography of the substrates and, in particular, the lip or rim of the pores. A previous study with these types of films⁴¹ highlighted that proteinaceous deposits selectively accumulate in these regions, and therefore in our cultures, it is likely that concentrations of collagen and other ECM molecules at the pore rim would form foci for mineralization. This organization of proteins and glycosaminoglycans on the membrane surface is perhaps dictated by a combination of altered surface charge and/or nanostructure, which influences the way biomolecules and water interface with the surface.^42–44 Further organization of the extracellular matrix is then cell-dependent based upon the structural contact guidance cues.

The spherical deposits of the mineralized matrix were much smaller (0.5–1 μm in diameter) and could be found at the surface, but also particularly inside the pores of the PCL film (Fig. 5). These type of structures have been described previously in vivo and are believed to be a functional unit of mineralization that combines protein, glycosaminoglycans, lipoproteins, and calcium and phosphate ions.^45–50 Often referred to as “calcospherites” or “calcospherulites,” these mineralized deposits are typically found in regions of rapid matrix deposition and maturation and are particularly localized to the mineralization front, where they are believed to be a part of the mechanism that leads to a uniform distribution of the mineral phase. EDX analysis revealed a Ca:P ratio of 1.3 in these deposits, which perhaps represents the relative immaturity of these mineralized structures.^49,50

To try and gain mechanistic insight into the differential levels of mineralized matrix deposition observed on the different surfaces investigated in this study, we investigated the size, abundance, and organization of vinculin-positive focal adhesion complexes formed in the cells. Our data show that a number of focal adhesion complexes (1–5 μm² in size) per unit cell area were significantly reduced on surfaces with 5–6, 10–11, and 15–16-μm pores (Fig. 6) and that the relative abundance of smaller focal adhesion complexes is increased in cells on 3–4 and 5–6 μm porosity PCL (Fig. 7). However, these observations do not absolutely correlate with the patterns of the mineralized matrix observed in this study suggesting that substrate-induced changes in cell shape, adhesion structure, and regulation of cell phenotype may not exclusively contribute to the patterns of matrix deposition.

Detailed CSLM revealed that the osteoblasts cultured on these surfaces exhibited globular structures that extended into the porosity of the PCL film (Fig. 8). This was particularly observed with the films that had 3–4 and 5–6-μm pores correlating with smaller vinculin-positive structures. It may be that on these surfaces, weaker cytoskeletal tensioning as a result of smaller vinculin containing anchoring points is more permissive for membrane extension and formation of membrane protrusions. Extensions of the plasma membrane have been previously described as podosomes or invadapodia with functions that range from bone resorption in osteoclasts to tissue invasion in tumor cells.^51,52 The common function of these structures is their use to create an extracellular space in which the molecular environment (including pH, balance of proteases/inhibitors) is carefully controlled. Our studies lead us to hypothesize that in an analogous way the PCL films with particular pore dimensions allow osteoblasts to create an extracellular space in which, the conditions particularly favor the deposition and formation of mineralized matrix structures.

Conclusions

These data demonstrate that regular porous films with pore dimensions in the order 3–4 and 5–6 μm support differential osteoblast adhesion. In particular, these surfaces control cell morphology and cell:ECM interactions, which in turn result in enhanced osteoblast deposition of the extracellular matrix and its mineralization. These surfaces provide a reproducible tool to dissect the mechanistic basis for substrate control of cell function and may also provide the basis for new therapeutic implants.

Footnotes

Acknowledgments

This work is supported by Grants-in-Aid and Special Coordination Funds for Promoting Science and Technology of Ministry of Education, Culture, Sports, Science and Technology, Japan and the Nuffield Foundation in the UK.

Disclosure Statement

The authors affirm that they have no conflict of interests in relation to the work presented in this manuscript.

References

Brunette

D.M.

, Kenner

G.S.

, Gould

T.R.

Grooved titanium surfaces orient growth and migration of cells from human gingival explants. J Dent Res, 62:1045. 1983.

Clark

, Connolly

, Curtis

A.S.

, Dow

J.A.

, Wilkinson

C.D.

Topographical control of cell behaviour. I. Simple step cues. Development, 99:439. 1987.

Clark

, Connolly

, Curtis

A.S.

, Dow

J.A.

, Wilkinson

C.D.

Topographical control of cell behaviour: II. Multiple grooved substrata. Development, 108:635. 1990.

Clark

, Connolly

, Curtis

A.S.

, Dow

J.A.

, Wilkinson

C.D.

Cell guidance by ultrafine topography in vitro. J Cell Sci, 99,Pt 1:73. 1991.

, Chehroudi

, Brunette

D.M.

The use of micromachined surfaces to investigate the cell behavioural factors essential to osseointegration. Oral Dis, 2:102. 1996.

Wojciak-Stothard

, Curtis

A.S.

, Monaghan

, McGrath

, Sommer

, Wilkinson

C.D.

Role of the cytoskeleton in the reaction of fibroblasts to multiple grooved substrata. Cell Motil Cytoskeleton, 31:147. 1995.

Oakley

, Jaeger

N.A.

, Brunette

D.M.

Sensitivity of fibroblasts and their cytoskeletons to substratum topographies: topographic guidance and topographic compensation by micromachined grooves of different dimensions. Exp Cell Res, 234:413. 1997.

Raz

, Lohmann

C.H.

, Turner

, Wang

, Poythress

, Blanchard

, Boyan

B.D.

, Schwartz

1alpha,25(OH)2D3 regulation of integrin expression is substrate dependent. J Biomed Mater Res Part A, 71:217. 2004.

Dalby

M.J.

, McCloy

, Robertson

, Agheli

, Sutherland

, Affrossman

, Oreffo

R.O.

Osteoprogenitor response to semi-ordered and random nanotopographies. Biomaterials, 27:2980. 2006.

10.

Dalby

M.J.

, McCloy

, Robertson

, Wilkinson

C.D.

, Oreffo

R.O.

Osteoprogenitor response to defined topographies with nanoscale depths. Biomaterials, 27:1306. 2006.

11.

Hamilton

D.W.

, Brunette

D.M.

The effect of substratum topography on osteoblast adhesion mediated signal transduction and phosphorylation. Biomaterials, 28:1806. 2007.

12.

Dalby

M.J.

Topographically induced direct cell mechanotransduction. Med Eng Phys, 27:730. 2005.

13.

Cai

, Bossert

, Jandt

K.D.

Does the nanometre scale topography of titanium influence protein adsorption and cell proliferation?

Colloids Surf B Biointerfaces, 49:136. 2006.

14.

Cooper

L.F.

A role for surface topography in creating and maintaining bone at titanium endosseous implants. J Prosthet Dent, 84:522. 2000.

15.

Masuda

, Yliheikkila

P.K.

, Felton

D.A.

, Cooper

L.F.

Generalizations regarding the process and phenomenon of osseointegration. Part I. In vivo studies. Int J Oral Maxillofac Implants, 13:17. 1998.

16.

Pilliar

R.M.

Cementless implant fixation—toward improved reliability. Orthop Clin North Am, 36:113. 2005.

17.

Wong

, Eulenberger

, Schenk

, Hunziker

Effect of surface topology on the osseointegration of implant materials in trabecular bone. J Biomed Mater Res, 29:1567. 1995.

18.

Neve

, Corrado

, Cantatore

F.P.

Osteoblast physiology in normal and pathological conditions. Cell Tissue Res, 343:289. 2011.

19.

Anderson

H.C.

Mineralization by matrix vesicles. Scan Electron Microsc, 953. 1984.

20.

Dalby

M.J.

, Di Silvio

, Gurav

, Annaz

, Kayser

M.V.

, Bonfield

Optimizing HAPEX topography influences osteoblast response. Tissue Eng, 8:453. 2002.

21.

Degasne

, Basle

M.F.

, Demais

, Hure

, Lesourd

, Grolleau

, Mercier

, Chappard

Effects of roughness, fibronectin and vitronectin on attachment, spreading, and proliferation of human osteoblast-like cells (Saos-2) on titanium surfaces. Calcif Tissue Int, 64:499. 1999.

22.

Martin

J.Y.

, Schwartz

, Hummert

T.W.

, Schraub

D.M.

, Simpson

, Lankford

Jr. , Dean

D.D.

, Cochran

D.L.

, Boyan

B.D.

Effect of titanium surface roughness on proliferation, differentiation, and protein synthesis of human osteoblast-like cells (MG63)

J Biomed Mater Res, 29:389. 1995.

23.

Anselme

, Bigerelle

, Noel

, Dufresne

, Judas

, Iost

, Hardouin

Qualitative and quantitative study of human osteoblast adhesion on materials with various surface roughnesses. J Biomed Mater Res, 49:155. 2000.

24.

Dalby

M.J.

, Gadegaard

, Tare

, Andar

, Riehle

M.O.

, Herzyk

, Wilkinson

C.D.

, Oreffo

R.O.

The control of human mesenchymal cell differentiation using nanoscale symmetry and disorder. Nat Mater, 6:997. 2007.

25.

Lohmann

C.H.

, Tandy

E.M.

, Sylvia

V.L.

, Hell-Vocke

A.K.

, Cochran

D.L.

, Dean

D.D.

, Boyan

B.D.

, Schwartz

Response of normal female human osteoblasts (NHOst) to 17beta-estradiol is modulated by implant surface morphology. J Biomed Mater Res, 62:204. 2002.

26.

Boyan

B.D.

, Bonewald

L.F.

, Paschalis

E.P.

, Lohmann

C.H.

, Rosser

, Cochran

D.L.

, Dean

D.D.

, Schwartz

, Boskey

A.L.

Osteoblast-mediated mineral deposition in culture is dependent on surface microtopography. Calcif Tissue Int, 71:519. 2002.

27.

Nemoto

, Uraki

, Kishimoto

, Sano

, Funada

, Obata

, Yabu

, Tanaka

, Shimomura

Production of mesoscopically patterned cellulose film. Bioresour Technol, 96:1955. 2005.

28.

Nishikawa

, Ookura

, Nishida

, Arai

, Hayashi

, Kurono

, Sawadaishi

, Hara

, Shimomura

Fabrication of honeycomb film of an amphiphilic copolymer at the air-water interface. Langmuir, 18:5734. 2002.

29.

Sato

, Hasebe

, Tanaka

, Takebayashi

, Nishikawa

, Shimomura

, Kawai

, Matsushita

, Todo

Preparation of the honeycomb patterned porous film of biodegradable polymer for tissue engineering scaffolds. Int J Nanosci Ser, 1:689. 2002.

30.

Tanaka

, Takebayashi

, Miyama

, Nishida

, Shimomura

Design of novel biointerfaces (II). Fabrication of self-organized porous polymer film with highly uniform pores. Bio-Med Mater Eng, 14:439. 2004.

31.

Yabu

, Takebayashi

, Tanaka

, Shimomura

Superhydrophobic and lipophobic properties of self-organized honeycomb and pincushion structures. Langmuir, 21:3235. 2005.

32.

Yabu

, Tanaka

, Ijiro

, Shimomura

Preparation of honeycomb-patterned polyimide films by self-organization. Langmuir, 19:6297. 2003.

33.

Tanaka

, Takayama

, Ito

, Sunami

, Yamamoto

, Shimomura

Effect of pore size of self-organized honeycomb-patterned polymer films on spreading, focal adhesion, proliferation, and function of endothelial cells. J Nanosci Nanotechnol, 7:763. 2007.

34.

Nishimura

, Yamada

Transfer of ganglioside GM3 oligosaccharide from a water soluble polymer to ceramide by ceramide glycanase. A novel approach for the chemical-enzymatic synthesis of glycosphingolipids. J Am Chem Soc, 119:10555. 1997.

35.

Bokhari

M.A.

, Akay

, Zhang

, Birch

M.A.

The enhancement of osteoblast growth and differentiation in vitro on a peptide hydrogel-polyHIPE polymer hybrid material. Biomaterials, 26:5198. 2005.

36.

Biggs

M.J.

, Richards

R.G.

, Gadegaard

, McMurray

R.J.

, Affrossman

, Wilkinson

C.D.

, Oreffo

R.O.

, Dalby

M.J.

Interactions with nanoscale topography: adhesion quantification and signal transduction in cells of osteogenic and multipotent lineage. J Biomed Mater Res Part A, 91:195. 2009.

37.

Khang

, Choi

, Im

Y.M.

, Kim

Y.J.

, Jang

J.H.

, Kang

S.S.

, Nam

T.H.

, Song

, Park

J.W.

Role of subnano-, nano- and submicron-surface features on osteoblast differentiation of bone marrow mesenchymal stem cells. Biomaterials, 33:5997. 2012.

38.

Holle

A.W.

, Engler

A.J.

More than a feeling: discovering, understanding, and influencing mechanosensing pathways. Curr Opin Biotechnol, 22:648. 2011.

39.

Dalby

M.J.

, Biggs

M.J.

, Gadegaard

, Kalna

, Wilkinson

C.D.

, Curtis

A.S.

Nanotopographical stimulation of mechanotransduction and changes in interphase centromere positioning. J Cell Biochem, 100:326. 2007.

40.

Tsimbouri

P.M.

, Murawski

, Hamilton

, Herzyk

, Oreffo

R.O.

, Gadegaard

, Dalby

M.J.

A genomics approach in determining nanotopographical effects on MSC phenotype. Biomaterials, 34:2177. 2013.

41.

Yamamoto

, Tanaka

, Sunami

, Ito

, Yamashita

, Morita

, Shimomura

Effect of honeycomb-patterned surface topography on the adhesion and signal transduction of porcine aortic endothelial cells. Langmuir, 23:8114. 2007.

42.

Tanaka

, Mochizuki

Effect of water structure on blood compatibility—thermal analysis of water in poly(meth)acrylate. J Biomed Mater Res Part A, 68:684. 2004.

43.

Tanaka

, Mochizuki

Clarification of the blood compatibility mechanism by controlling the water structure at the blood-poly(meth)acrylate interface. J Biomater Sci Polym Ed, 21:1849. 2010.

44.

Tanaka

, Mochizuki

, Ishii

, Motomura

, Hatakeyama

Study of blood compatibility with poly(2-methoxyethyl acrylate). Relationship between water structure and platelet compatibility in poly(2-methoxyethylacrylate-co-2-hydroxyethylmethacrylate)

Biomacromolecules, 3:36. 2002.

45.

Aaron

J.E.

, Oliver

, Clarke

, Carter

D.H.

Calcified microspheres as biological entities and their isolation from bone. Histochem J, 31:455. 1999.

46.

Boyde

, Sela

Scanning electron microscope study of separated calcospherites from the matrices of different mineralizing systems. Calcif Tissue Res, 26:47. 1978.

47.

Ornoy

, Atkin

, Levy

Ultrastructural studies on the origin and structure of matrix vesicles in bone of young rats. Acta Anat, 106:450. 1980.

48.

Carter

D.H.

, Scully

A.J.

, Davies

R.M.

, Aaron

J.E.

Evidence for phosphoprotein microspheres in bone. Histochem J, 30:677. 1998.

49.

Midura

R.J.

, Vasanji

, Su

, Midura

S.B.

, Gorski

J.P.

Isolation of calcospherulites from the mineralization front of bone. Cells Tissues Organs, 189:75. 2009.

50.

Midura

R.J.

, Vasanji

, Su

, Wang

, Midura

S.B.

, Gorski

J.P.

Calcospherulites isolated from the mineralization front of bone induce the mineralization of type I collagen. Bone, 41:1005. 2007.

51.

Saltel

, Daubon

, Juin

, Ganuza

I.E.

, Veillat

, Genot

Invadosomes: intriguing structures with promise. Eur J Cell Biol, 90:100. 2011.

52.

Linder

The matrix corroded: podosomes and invadopodia in extracellular matrix degradation. Trends Cell Biology, 17:107. 2007.