Sage Journals: Discover world-class research

Abstract

Through mediation of integrin clustering, nanoscale surface structure of carbon nanotubes (CNTs) directly affects cell binding and subsequent behavior. The influence of morphological structures on proliferation, differentiation, and organization of focal adhesions and cytoskeleton of human mesenchymal stem cells (hMSCs) was studied. The following two surface morphologies were fabricated and examined: a random network multiwalled carbon nanotubes film (RNCNT) and a vertically aligned multiwalled carbon nanotube (VACNT) film. hMSCs adhered and spread earlier on both CNT surfaces than on control. A statistically significantly increased number of attached cells were observed on VACNT surfaces. No CNT substrate enhanced differentiation, but both maintained the differentiation property of hMSCs. VACNTs recruited vinculin from the cytoplasm for the formation of focal adhesion complexes earlier in comparison to RNCNT. There is still disparity on how nanostructures regulate the progression toward an osteoblastic phenotype. It is necessary to explore various architectures in order to understand how they initiate osteoinductive or proliferating signals.

Keywords

carbon nanotube topography human mesenchymal cells focal adhesion

Introduction

Numerous chemical, physical, mechanical, and electrical stimuli are integral for tissue regeneration. Moreover, nanotopography, which mimics the local nanoscale topographic cues within the stem cell niche, regulates multipotent cell behavior.

The stem cell/material interface is a highly complex, dynamic microenvironment, and both cell and material cooperate to its creation. Cells sense substrate material properties, such as matrix structure, elasticity, and composition, integrate cues via signal propagation and, ultimately, translate parallel signaling information into cell fate direction control.^1,2

Stem cell differentiation is influenced by cell shape. The cell spreading area and contour morphology have been considered as good indicators of the interaction with surfaces. However, the underlined processes by which cytoskeleton assembly and architecture can be regulated by interaction at the cell-material interface are not fully understood. Synthetic nanostructures can drive specific cell differentiation, but the sensing mechanism of nanocues remains poorly understood.³

Research has been performed to understand whether topographical symmetry and disorder can control cell response. Nanoscale disorder has been shown to stimulate human mesenchymal stem cells (hMSCs) to produce bone mineral in vitro in the absence of osteogenic supplements.⁴ Regular topography reduces cell adhesion very markedly compared with less regular arrays or planar surfaces. The cells appear to be able to distinguish between different symmetries of array.^5,6 Moreover, osteoblast-like cells have been found that can recognize surface nanopatterns with vertical variations in the subnanometric range, which significantly influences their response.⁷

Studies on the interaction of cells with carbon nanotubes (CNTs) have been receiving increasing attention owing to their potential for various cellular applications. Purified pristine multiwalled carbon nanotubes (MWCNTs) have been found to support adhesion and proliferation of hMSCs, enhance mature osteogenic gene expression, and accelerate cell differentiation, even in the absence of osteogenic supplements.^8,9 Controlling thickness, roughness, and directional alignment of single-walled CNT films would regulate the growth and differentiation ofstem cells.¹⁰ On arrays of vertically aligned MWCNTs (VACNTs), unlike cell growth on nonpatterned surfaces, the osteoblast-like cell attachment and spreading process on VACNT was significantly enhanced. The periodicity and alignment of VACNT considerably influenced growth, shape, and orientation of the cells.¹¹

It has been suggested that nanotopography modulates cell behavior by changing the integrin clustering and focal adhesion (FA) assembly, leading to changes in cytoskeletal organization and cell mechanical properties.^12,13 FAs are known to be involved in the process of specific pattern recognition and subsequent response by cells.¹⁴

In this study, the influence of morphological structures of MWCNT films on cellular behavior, cell proliferation, and differentiation was examined. In addition, the study was particularly focused on the expression and organization of FAs and cytoskeleton of hMSCs, in the direction of discovering the mechanisms by which cells respond to surface nanotexturing. The following two surface morphologies were examined: a random network MWCNT (RNCNT) film and a VACNT film. Determining the cell-material interaction at the interface will ultimately lead toward engineering of cell-instructive materials.

Materials and Methods

Materials

The pristine MWCNTs used in this study were obtained from Nanothinx S.A. (purity 98.3%). MWCNTs were in black powder form and synthesized by chemical vapor deposition method. The mean diameter and length of MWCNTs were reported by the company to be 5–20 nm and larger than 10 µm, respectively.

MWCNT film preparation

Films of the following two surface morphologies were examined: a random network MWCNT film and a vertically aligned MWCNT film. In a 250–mL round bottom flask, pristine MWCNT (20 mg), sodium dodecyl sulfate (SDS) surfactant (5 mg), and deionized water (50 mL) were added, and the mixture was treated in an ultrasonic bath for half an hour. SDS (L4509; Sigma-Aldrich) was added in the mixture in order to improve the dispersion of MWCNT in the water. The initial dispersion was facilitated by ultrasound sonication of the samples for four hours inside an ultrasound bath. Remaining aggregates in the suspension were removed by centrifugation at 1800 × g for four minutes. The mixture was filtrated through a 200-nm nylon membrane, and the obtained film was washed several times with deionized water (2 L). Sterilization was performed at 121°C for 15 minutes in an autoclave. VACNT films were prepared and characterized by Nanothinx S.A.

Characterization of MWCNT films

Surface morphology was assessed using the scanning electron microscopy (SEM; JEOL-JSM 6300). Surface roughness was assessed using the atomic force microscopy (AFM; Nanoscope IIIa) in tapping mode on 10 µm × 10 µm areas. The surface roughness parameter R_α (roughness average) was calculated. Results were presented as mean ± standard deviation (SD) of three experiments (six areas per sample and three samples of each substrate on total were examined).

Cell culture

hMSCs, obtained by aspiration from the femoral diaphysis of healthy male and female patients aged 40–70 years undergoing hip surgery after trauma, were used for these experiments. A single-cell suspension was prepared by repeatedly aspirating the cells successively through 19- and 21-G needles. The characterization of the phenotype of hMSCs was performed by direct flow cytometric analysis of the expression of the following specific mesenchymal surface antigens: CDw90 (CD90-phycoerythrin cyanine 5/PC5; PN IM3703; ImmunoTech, Beckman Coulter), CD105 (CD105-phycoerythrin/PE; PN A07414; ImmunoTech, Beckman Coulter), and CD29 (CD29-fluorescein isothiocyanate/FITC; 0791; ImmunoTech, Beckman Coulter).

The cell suspension was cultured in osteogenic medium (Alpha Medium [F0925; Biochrom] with 10% fetal bovine serum (FBS) [S0615; Biochrom], 2 mM l-glu-tamine [K0283; Biochrom], 50 mg/mL l-ascorbic acid [A4034; Sigma-Aldrich], 50 µm/mL gentamicin [A2712; Biochrom], 2.5 µg/mL amphotericin β [A2610; Biochrom], 10^-8 M dexamethasone [Sigma-Aldrich], and 10 mM Na-β-glycerophosphate [G9422; Sigma-Aldrich]).

The cells were maintained in humidified incubator at 37°C with 5% CO₂ and 95% air. A medium replacement was carried out every three to four days. Cell culture on plastic surface (TCP) was used as control. hMSCs were expanded in vitro until passage 6 and, consequently, seeded onto the MWCNT films at subconfluence density (3000 cells/cm²). Previously, samples had been placed into 24-well plates and preconditioned with Alpha Medium for one hour at 37°C.

MWCNT films toxicity

The toxicity of the MWCNT films on hMSCs was determined using the Pierce LDH Cytotoxicity Assay Kit (Thermo Fisher Scientific), which is a reliable colorimetric assay to quantitatively measure lactate dehydrogenase (LDH) released into the supernatant medium from damaged cells as a biomarker for cellular cytotoxicity and cytolysis, using an enzymatic reaction that results in a red formazan product, which can be measured spectrophotometrically at 490 nm. The cytotoxicity value for each substrate was determined as the ratio of the released in the medium LDH to the total LDH (cellular + extracellular). As a positive control, the cells have been lysed before measurement using the ToxiLight Lysis Reagent (Lonza Cologne). For negative control, medium + fetal bovine serum was used.

Cell proliferation

The cells were incubated at 37°C in a CO₂ cell culture incubator. After one, three, and seven days of incubation, the cells were treated with 0.1% of Triton-X 100 and gently shaken for 10 minutes on an orbital shaker at room temperature. The standard protocol for a commercially available colorimetric proliferation assay was used to quantify cell proliferation (Pierce LDH Cytotoxicity Assay Kit; Thermo Fisher Scientific).

Alkaline phosphatase activity

Alkaline phosphatase (ALP) enzyme activity of hMSCs, either cultured on CNTs films or cultured on control (TCP), was assessed after one, three, and seven days of cell culture. ALP activity from the cell samples was quantified by the specific conversion of p-nitrophenyl phosphate (P7998; Sigma-Aldrich) into p-nitrophenol. Cells were rinsed three times in Phosphate Buffer Solution (PBS) (L1835; Biochrom) and lysed by 0.1% Triton X-100 (X100; Sigma-Aldrich) in PBS. A total of 100 µL of lysate was transferred into 96-well plate, and 100 µL of reaction buffer (0.1 M glycine, pH 10.4 [G7126; Sigma-Aldrich], 1 mM MgCl₂ [M0250; Sigma-Aldrich], 1 mM ZnCl₂ [20,808-6; Sigma-Aldrich], and 1 mg/mL p-nitrophenyl phosphate [S0942; Sigma-Aldrich]) was added and incubated for 45 minutes at room temperature. The reaction was stopped by adding 50 µL of 3 M NaOH and read spectrophotometrically at 405 nm. The results were displayed as ALP activity per milligram of intracellular total protein content (quantified using the Total Protein Kit [TP0400; Sigma-Aldrich]).

Immunocytochemistry for cytoskeletal organization and FA quantification

Cells were cultured on the materials for one, three, and seven days, washed in PBS, fixed in 4% formaldehyde in PBS, permeabilized in 1% Triton X-100 in PBS, and incubated with PBS containing 5% BSA to block nonspecific epitopes. The Actin Cytoskeleton and Focal Adhesion Staining Kit (FAK100; Millipore) was used for cytoskeletal organization and FA visualization.¹⁵ The kit contains a fluorescent-labeled phalloidin (TRITC-conjugated phalloidin) to map the local orientation of actin filaments within cell and a monoclonal antibody specific for vinculin (which is a part of FA contacts), followed by goat antimouse secondary antibody Alexa Fluor 488. 4′,6-diamidino-2-phenylindole (DAPI) was used for the fluorescent labeling of nuclei.

Later, the cells were mounted by using antifade mounting solution (Millipore 5013). The samples were observed with an inverted confocal microscope (Nikon D-eclipse C1, ECLIPSE TE-2000U). For the quantification of FA, highresolution images of cells were processed through the software ImageJ 150c, which identifies the vinculin-stained areas (individual adhesions) as a percentage of the area of the whole cell, based on their size and intensity.

Statistical analysis

Data are from one of two replicate experiments and are presented as mean ± SD for three cultures. The data were analyzed using SPSS 14.0 software (SPSS). The level of significance was calculated using Student's t-test for single comparisons and ANOVA for multiple comparisons using a Tukey mean comparison test with 95% confidence level. Statistical significance was assumed at the 95% confidence limit or greater (P < 0.05).

Ethical considerations

This research was exempt from the requirement to obtain ethics committee approval because the used material was intended to be discarded and patients gave their written, informed consent to participate in the research. The research was conducted in accordance with the principles of the Declaration of Helsinki.

Results

Surface morphology

The morphology of the substrates, characterized by SEM and AFM, is shown in Figures 1 and 2, respectively. The R_α roughness values of the films, obtained via AFM scanning (tapping mode), of an area of 10 µm × 10 µm were 148.4 ± 21.3 and 67.4 ± 11.2 nm for RNCNTs and VACNTs, respectively. The roughness values of the two substrates were statistically different (P < 0.001).

Figure 1

SEM images of the topographies used are shown in (A) VACNTs and (B) RNCNTs.

Figure 2

AFM images of the topographies used are shown in (A) VACNTs and (B) RNCNTs.

MWCNT films toxicity

The toxicity levels for the hMSCs on both MWCNT substrates were not significantly different than those of the negative control at any culture time tested (P < 0.05) (Fig. 3). Statistically significantly lower values of toxicity (P < 0.001) were found when the toxicity levels were compared with the positive control (100% lysed cells). These results suggest that no significant membrane disruption occurred in the cells cultured on any MWCNT substrate.

Figure 3

Substrates cytotoxicity expressed as the ratio of the LDH released from damaged cells in the medium to the total LDH (cellular + extracellular). Lysed cells were used for the positive control values, and medium supplemented with FCS was used as the negative control (mean ± SD, n = 3). No statistical differences were found among substrates VACNTs and RNCNTs and negative control regarding cytotoxicity, whereas both substrates’ cytotoxicity was significantly lower than positive controls (P < 0.001).

Cell proliferation

The results of cell proliferation for one, three, and seven days are shown in Figure 4. On both MWCNT surfaces, the cells proliferated, and a statistically significantly (P < 0.01) increased number of cells at all times tested were observed in relation to control (TCP). On VACNTs, the proliferation, particularly after seven days of culture, was much more pronounced in comparison to RNCNTs (P < 0.01).

Figure 4

Cell proliferation for one, three, and seven days of culture (mean ± SD, n = 3). Asterisks indicate significant differences between columns connected with brackets, P < 0.01.

ALP activity

Figure 5 shows the results of ALP activity per milligram of intracellular total protein. ALP activity of the cells, cultured on both MWCNT substrates, was comparable to control at all culture time points of one, three, and seven days. This result suggests that MWCNTs do not enhance differentiation more than TCP but maintain the differentiation property of hMSCs. On VACNTs, ALP activity was higher but not significantly different than that on RNCNTs.

Figure 5

ALP activity per milligram of intracellular total protein for one, three, and seven days of culture (mean ± SD, n = 3). No statistical differences were found between VACNT and RNCNT substrates or between any MWCNT substrate and control.

Cytoskeletal organization and FA quantification

Cells spread earlier on both MWCNT substrates than on TCP (Fig. 6). As both MWCNTs’ surface patterning is isotropic, similar cytoskeletal structure was observed on these substrates: cells are cuboidal and not elongated. There is a clear difference in the spreading of cells on MWCNT with more filopodia or lamellipodia compared to TCP, particularly after seven days of culture, when cultures approach confluence (Fig. 6).

Figure 6

Cytoskeleton and distribution of focal adhesions visualized by immunofluorescence staining of vinculin (green), F-actin (red), and nuclei (blue) for one, three, and seven days of cell growth on the substrates. Bar = 50 ᴠm.

The FAs and their distribution, visualized by immunofluorescence staining of vinculin, can be observed at both the central region and the peripheral region and also at the end of the F-actin fiber bundles in the filopodia or lamellipodia. In contrast, a higher density of FAs was observed at the poles of the more elongated, and without noticeable protrusions, hMSCs cultured on TCP. Few FAs were present at the central region of the cells on TCP.

The vinculin level expressed by hMSCs was increased on both MWCNTs compared to control (Fig. 7) at all culture times tested. Moreover, a statistically significant increase in FAs was found on VACNTs in comparison to the RNCNTs (P < 0.05) on the first day of culture. On the contrary, on the third and seventh days, the observed increase in FAs was not statistically significant.

Figure 7

Vinculin expression level, as a percentage of the area of the whole cell, for one, three, and seven days of cell growth on each substrate (mean ± SD, n = 3). Asterisks indicate significant differences between columns connected with brackets, *P < 0.05 and **P < 0.01.

Discussion

Nanostructured materials are believed to play a fundamental role in orthopedic research because bone has a structural hierarchy in the nanometer regime. Recent developments in advanced micro- and nanofabrication techniques have enabled the fabrication of substrates that are able to reproduce the structure and length scale of bone native topography in two-dimensional substrates. Cells respond to two-dimensional synthetic topographic substrates in a wide range of responses, which depend upon cell type, feature size and geometry, and substrate stiffness.^3,16 The exact mechanism by which nanotextures can control cellular behavior is not well understood. Thus, it is unknown how cells recognize and respond to certain surface patterns, whereas a directed response appears to be absent on other pattern types. In this study, both examined MWCNT substrates of identical chemistry were isotropic but displayed altered nanoscale surface roughness. In addition, the vertically aligned CNTs, although neither patterned nor highly ordered, displayed more regular and ordered surface nanostructure with discrete nanotube shape in comparison to the random network MWCNT films.

In a previous study of the same group,⁸ it has been found that differentiation of hMSCs, cultured on RNCNTs, occurred even in the absence of additional biochemical-inducing agents, being as about fourfold as that for the control, whereas proliferation was delayed. When culturing with osteogenic supplements, the ALP expression on MWCNT was only 0.7-fold higher in comparison to control. In this study, no significant difference was observed between any MWCNT substrate and control. In addition, the cell proliferation on MWCNT substrates was enhanced in comparison to control. For these two studies, cells from the same origin were used, with only one difference: in this particular work, the cells were of the sixth to seventh passages, whereas, in Ref. 8, the cells were of the second to third passages. Studies have shown that increasing age and number of passages have lineage-dependent effects on mouse MSC differentiation potential,¹⁷ and in vitro expansion studies showed that hMSCs can be effectively expanded up to four passages (approximately 10–12 population doublings) to retain their multipotency.¹⁸ Working with hMSCs, Banfi et al¹⁹ found decreased adipogenic, chondrogenic, and osteogenic potentials when increasing from the first to fifth passages. Lohmann et al²⁰ indicated that surface roughness promotes osteogenic differentiation of less mature cells. As cells become more mature, they exhibit a reduced sensitivity to their substrate, but even the terminally differentiated osteocyte is affected by changes in surface roughness.

Expanding this result by including all above findings, it could be concluded that nanostructured substrates and, particularly, MWCNTs influence hMSC response according to their differentiation potential. When the osteogenic potential is decreased with the increasing passage number, MWCNTs and, particularly, VACNTs enhance the proliferation of hMSCs, at the expense of their differentiation.

Two important goals in stem cell research are to manage to control the cell proliferation without differentiation and to direct the differentiation into a specific cell lineage when desired. Current research explores such paths.²¹ hMSCs are attachment-dependent cells that interact with their substrate via cytoskeletal alterations. The interaction of bone cells with biomaterials influences both their proliferation and differentiation, through cytoskeletal-mediated mechanisms. Oh et al,²¹ using a range of nanotube diameters, demonstrated that a guided osteogenic differentiation of hMSCs can be controllably manipulated by selective sizing of the nanotube dimensions. Small (30 nm diameter) TiO₂ nanotubes promoted adhesion without noticeable differentiation, whereas larger (70–100 nm diameter) nanotubes elicited cell elongation, which induced differentiation into osteoblast-like cells. Similar results were found in this study with CNT: 5–20 nm diameter nanotubes supported proliferation, without inducing differentiation. Cells of the second to third passages or cultured without osteogenic supplements could be used to examine whether hMSCs’ maturation state would alter these results.

VACNTs, whose patterning resembles nanopit array, support the proliferation of hMSCs much more than RNCNTs. In addition, VACNTs enhanced differentiation in comparison to RNCNT networks and control but not statistically significantly. Nanopits generally reduce initial osteoblast cell attachment in respect to the nonpatterned substrate with same chemistry. No obvious trends to predict the effect of nanopit geometries on proliferation exist in current literature. Some combinations of geometry, length scale, substrate material, and cell types promote more rapid proliferation, while others reduce proliferation rate.¹⁶ The symmetry and order of the nanopits were found to significantly affect the differentiation of hMSCs. While hMSCs cultured on completely ordered or completely random nanopits did not lead to expression of two bone-specific extracellular matrix (ECM) proteins, osteopontin and osteocalcin, hMSCs cultured on slightly irregular substrates did exhibit significant amounts of these proteins of interest. The results of this study, culturing hBMCs on slightly ordered nanopatterned substrates, generally support the above findings.

VACNTs recruit earlier vinculin from the cytoplasm for the formation of FA complexes in comparison to RNCNTs. FAs are known to be involved in the process of specific pattern recognition and subsequent response by cells.²² The effects of pit diameter on FA formation were recently demonstrated²³ using TiO₂ nanotubes. It was shown experimentally that a nanotube diameter smaller than 30 nm, with a maximum at 15 nm, provided an effective length scale for accelerated integrin clustering and FA formation, and that this length scale strongly enhanced cellular activities of MSCs compared to smooth TiO₂ surfaces. We have similarly demonstrated results that are in agreement with these findings: increased formation of FA was observed on VACNTs with nanotube diameter smaller than 30 nm. In addition, Lim et al²⁴ concluded that greater cell adhesion and increased integrin expression occur when topographic features have less than 10–20 nm scale z-axis dimension (height or depth) and that this occurs despite topographic shapes (island or pit). This finding is consistent with the results of this study, where the less rough VACNT substrate supported the increased formation of FAs. The mechanisms by which nanotopographic cues influence hMSCs’ response appear to involve changes in cytoskeletal organization and structure in response to the substrate's patterns and features. The results of this study support the hypothesis that changes of the substrate's topography influence the clustering of adhesion molecules, thus altering the number and distribution of FAs.

Fibronectin and collagen type IV adsorption studies have shown a linear adsorption behavior as the surface nanoroughness increased on highly ordered poly(lactic-co-glycolic acid) surfaces of identical chemistry but altered nanoscale surface roughness and energy.²⁵ This result could explain the improved cellular responses on nanostructured surfaces commonly reported in the literature today. In this study, VACNTs, although display lower nanoroughness, support higher initial cell adhesion and proliferation. Thus, additional parameters should be taken into account, apart from initial protein adsorption events to explain the enhanced cellular adhesion, such as surface chemistry and cell differentiation potential.

VACNTs present a degree of toxicity, which is comparable to that of RNCNTs and negative control and much lower than the positive control. The vertical orientation of the CNTs on this particular substrate permits them to pierce more easily into cell membrane and disrupt it, resulting in an increased release of LDH in the supernatant medium from damaged cells. Simultaneously, proliferation was enhanced on the same substrate. Tutak et al²⁶ proposed a mechanism that can be adopted to interpret the current results: initially, VACNTs are absorbed in a process that resembles endocytosis, inducing acute toxicity. Nanotube-mediated cell destruction, however, induces a release of endogenous factors that act to boost the activity of the surviving cells by stimulating the synthesis of extracellular matrix. Absolute amount of LDH released in the culture medium by VACNTs, because of the larger number of cells present on this substrate, was statistically higher in comparison to both RNCNTs and TCP (P < 0.01) (data not shown). This finding, along with the mechanism proposed by Tutak et al, which reconciles the apparent toxicity and benefits of CNT substrates for osteoblast cell proliferation and differentiation, might provide the explanation of the enhanced response of hMSCs on VACNTs comparing to RNCNTs and control.

In conclusion, modulation of integrin-mediated FA and related cell signaling by altering nanoscale substrate topography will have powerful applications in biomaterials science and tissue engineering. Systematic control of the roughness and directional alignment of CNT films would provide a strategy for controlling the growth and maintenance of the differentiation property of stem cells. However, the maturation state or the differentiation potential of the cells (or the number of passage) is a critical parameter to design nanotechnology-based routes for orthopedics-related hMSC treatments. Further work is required to explore various architectures to understand how they initiate osteoinductive or proliferating signals.

Author Contributions

Conceived and designed the experiments: DD and AK. Analyzed the data: AK. Wrote the first draft of the manuscript: DD. Contributed to the writing of the manuscript: DD and AK. Agree with manuscript results and conclusions: AK and DD. Jointly developed the structure and arguments for the paper: AK and DD. Made critical revisions and approved final version: AK and DD. All authors reviewed and approved of the final manuscript.

Footnotes

Acknowledgment

We are grateful to Dr. G. Michanetzis for the images and measurements of roughness with AFM.

References

Langer

, Borenstein

J.T.

, Bettinger

C.J.

Engineering substrate topography at the micro- and nanoscale to control cell function. Angew Chem Int Ed. 2009; 48: 5406–5415.

Zhang

, Gordon

, Qian

, Chen

Engineering nanoscale stem cell niche: direct stem cell behavior at cell-matrix interface. Adv Healthc Mater. 2015; 16: 1900–1914.

Teo

B.K.

, Wong

S.T.

, Kiat

L.C.

Nanotopography modulates mechano-transduction of stem cell and induces differentiation through focal adhesion kinase. ACS Nano. 2013; 7: 4785–4798.

Dalby

M.J.

, Gadegaard

, Tare

. The control of human mesenchymal cell differentiation using nanoscale symmetry and disorder. Nat Mater. 2007; 6(12): 997–1003.

Curtis

A.S.

, Casey

, Gallagher

J.O.

, Pasqui

, Wood

M.A.

, Wilkinson

C.D.

Substratum nanotopography and the adhesion of biological cells. Are symmetry or regularity of nanotopography important?

Biophys Chem. 2001; 94: 275–283.

Curtis

A.S.

, Gadegaard

, Dalby

M.J.

, Riehle

M.O.

, Wilkinson

C.D.

, Aitchison

Cells react to nanoscale order and symmetry in their surroundings. IEEE Trans Nanobioscience. 2004; 3: 61–65.

Wittenbrink

, Hausmann

, Schickle

. Low-aspect ratio nanopatterns on bioinert alumina influence the response and morphology of osteoblast-like cells. Biomaterials. 2015; 62: 58–65.

Kroustalli

, Kourkouli

, Deligianni

Cellular function and adhesion mechanisms of human bone marrow mesenchymal stem cells on multi-walled carbon nanotubes. Ann Biomed Eng. 2013; 41: 2655–2665.

Deligianni

MWCNTs enhance hBMSCs spreading but delay their proliferation in the direction of differentiation acceleration. Cell Adh Migr. 2014; 8: 487–492.

10.

Lee

J.H.

, Shim

, Choolakadavil Khalid

. Random networks of single-walled carbon nanotubes promote mesenchymal stem cell's proliferation and differentiation. ACS Appl Mater Interfaces. 2015; 28(7): 1560–1567.

11.

Giannona

, Firkowska

, Rojas-Chapana

, Giersig

Vertically aligned carbon nanotubes as cytocompatible material for enhanced adhesion and proliferation of osteoblast-like cells. J Nanosci Nanotechnol. 2007; 7: 1679–1683.

12.

Yim

E.K.

, Darling

E.M.

, Kulangara

, Guilak

, Leong

K.W.

Nanotopographyinduced changes in focal adhesions, cytoskeletal organization, and mechanical properties of human mesenchymal stem cells. Biomaterials. 2010; 31: 1299–1306.

13.

Biggs

M.J.

, Richards

R.G.

, Dalby

M.J.

Nanotopographical modification: a regulator of cellular function through focal adhesions. Nanomedicine. 2010; 6: 619–633.

14.

Lamers

, te Riet

, Domanski

. Dynamic cell adhesion and migration on nanoscale grooved substrates. Eur Cell Mater. 2012; 23: 182–193.

15.

Owen

G.R.

, Meredith

D.O.

, ap Gwynn

, Richards

R.G.

Focal adhesion quantification–a new assay of material biocompatibility? Review. Eur Cell Mater. 2005; 9: 85–96.

16.

Bettinger

C.J.

, Langer

, Borenstein

J.T.

Engineering substrate topography at the micro- and nanoscale to control cell function. Angew Chem Int Ed. 2009; 48: 5406–5415.

17.

Kretlow

J.D.

, Jin

Y-Q

, Liu

. Donor age and cell passage affects differentiation potential of murine bone marrow-derived stem cells donor variation and loss of multipotency during in vitro expansion of human mesenchymal stem cells for bone tissue engineering. BMC Cell Biol. 2008; 9: 60–73.

18.

Siddappa

, Licht

, van Blitterswijk

, de Boer

Donor variation and loss of multipotency during in vitro expansion of human mesenchymal stem cells for bone tissue engineering. J Orthop Res. 2007; 25: 1029–1041.

19.

Banfi

, Muraglia

, Dozin

, Mastrogiacomo

, Cancedda

, Quarto

Proliferation kinetics and differentiation potential of ex vivo expanded human bone marrow stromal cells: implications for their use in cell therapy. Exp Hematol. 2000; 28(6): 707–715.

20.

Lohmann

C.H.

, Bonewald

L.F.

, Sisk

M.A.

. Maturation state determines the response of osteogenic cells to surface roughness and 1,25-dihydroxyvitamin D3. J Bone Miner Res. 2000; 15: 1169–1180.

21.

, Brammer

K.S.

, Li

Y.S.

. Stem cell fate dictated solely by altered nanotube dimension. Proc Natl Acad Sci USA. 2009; 106: 2130–2135.

22.

Lamers

, van Horssen

, te Riet

. The influence of nanoscale topographical cues on initial osteoblast morphology and migration. Eur Cell Mater. 2010; 20: 329–343.

23.

Park

, Bauer

, von der Mark

, Schmuki

Nanosize and vitality: TiO2 nanotube diameter directs cell fate. Nano Lett. 2007; 7: 1686–1691.

24.

Lim

J.Y.

, Dreiss

A.D.

, Zhou

. The regulation of integrin-mediated osteoblast focal adhesion and focal adhesion kinase expression by nanoscale topography. Biomaterials. 2007; 28: 1787–1797.

25.

Ercan

, Khang

, Carpenter

, Webster

T.J.

Using mathematical models to understand the effect of nanoscale roughness on protein adsorption for improving medical devices. Int J Nanomedicine. 2013; 8(suppl 1): 75–81.

26.

Tutak

, Park

K.H.

, Vasilov

. Toxicity induced enhanced extracellular matrix production in osteoblastic cells cultured on single-walled carbon nanotube networks. Nanotechnology. 2009; 20: 255101.