Electrospun polyvinyl alcohol–collagen–hydroxyapatite nanofibers: a biomimetic extracellular matrix for osteoblastic cells This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2012 Nanotechnology 23 115101 (http://iopscience.iop.org/0957-4484/23/11/115101) Download details: IP Address: 141.217.203.141 The article was downloaded on 26/04/2012 at 19:15 Please note that terms and conditions apply. View the table of contents for this issue, or go to the journal homepage for more Home Search Collections Journals About Contact us My IOPscience
Transcript
Electrospun polyvinyl alcohol–collagen–hydroxyapatite nanofibers: a biomimetic extracellular
matrix for osteoblastic cells
This article has been downloaded from IOPscience. Please scroll down to see the full text article.
Electrospun polyvinylalcohol–collagen–hydroxyapatitenanofibers: a biomimetic extracellularmatrix for osteoblastic cellsWei Song1, David C Markel2,3, Sunxi Wang4, Tong Shi1,Guangzhao Mao4 and Weiping Ren1,2,5
1 Department of Biomedical Engineering, Wayne State University, Detroit, MI, USA2 Detroit Medical Center and Providence Hospital Orthopedic Residency, Detroit, MI, USA3 Department of Orthopedic Surgery, Providence Hospital, Southfield, MI, USA4 Department of Chemical Engineering and Materials Science, Wayne State University, Detroit, MI, USA
Received 18 November 2011, in final form 7 January 2012Published 28 February 2012Online at stacks.iop.org/Nano/23/115101
AbstractThe failure of prosthesis after total joint replacement is due to the lack of early implantosseointegration. In this study polyvinyl alcohol–collagen–hydroxyapatite (PVA-Col-HA)electrospun nanofibrous meshes were fabricated as a biomimetic bone-like extracellular matrixfor the modification of orthopedic prosthetic surfaces. In order to reinforce the PVAnanofibers, HA nanorods and Type I collagen were incorporated into the nanofibers. Weinvestigated the morphology, biodegradability, mechanical properties and biocompatibility ofthe prepared nanofibers. Our results showed these inorganic–organic blended nanofibers to bedegradable in vitro. The encapsulated nano-HA and collagen interacted with the PVA content,reinforcing the hydrolytic resistance and mechanical properties of nanofibers that providedlonger lasting stability. The encapsulated nano-HA and collagen also enhanced the adhesionand proliferation of murine bone cells (MC3T3) in vitro. We propose the PVA-Col-HAnanofibers might be promising modifying materials on implant surfaces for orthopedicapplications.
(Some figures may appear in colour only in the online journal)
1. Introduction
Each year, over 700 000 total joint replacement (TJR)operations (hip and knee) are performed in the UnitedStates [1, 2], and the numbers continue to increase withan aging population. While TJR improves the quality oflife for these patients, the number of revisions for failedimplants is rising due to the increasing pool of peoplewith implanted devices and longer life spans. A lack ofearly implant osseointegration represents one of the risk
5 Author to whom any correspondence should be addressed.
factors leading to implant instability, micromotion andosteolysis/loosening [3–5].
Natural bone is composed of a complex hierarchi-cal structure with mineralized collagen fibrils and HAnanocrystals [6]. The organic matrix is mainly Type Icollagen (Col) which provides bone with its flexibilityand resilience, and is important for cell adhesion andmigration [6]. The inorganic phase is composed of the mineralhydroxyapatite (HA) which is responsible for the stiffnessand strength of bone. Nano-scale-HA crystals are crucialfor osteoinduction, biomineralization and osteointegration [7].The organic–inorganic constituents combine together toprovide a mechanical and supportive role in the body [8].
Osseointegration is defined as the direct anchorage of animplant by bone formation at the bone–implant surface. Manyresearch efforts have been directed toward improving thebone–implant interface, with the aim of accelerating bonehealing and improving bone anchorage to the implant [9–15].One of many research strategies is to develop a biopolymer-based implant surface with a ‘bone-like’ nature (structural,mechanical and biological behavior) [16–19]. HA compositeswith biodegradable polymers, such as poly lactide (PLA) orpoly (lactide-co-glycolide) (PLGA) [20], have been fabricatedto reduce the fragility of the implant surface. However,unfavorable effects of acidic degradation products (lactic andglycolic acid) from these polymers on the cells surroundingthe implants have become a concern [21, 22]. Polyvinylalcohol (PVA), a water-soluble and biodegradable polymer,has been used extensively in the pharmaceutical industrybecause of its biocompatibility, proven mechanical strengthand anabolic effect on bone formation [23–26]. In addition,PVA has a self-crosslink capability (film or hydrogel forming)due to the abundant number of hydroxyl groups in its sidechains [23–26].
The nano-scale topography of an implant surface iscritical for the capacity of osseointegration [27–29]. Differenttopographies, methods and materials have been developed toenhance osseointegration [30–34]. One alternative approachhas been electrospun nanofibers [35–38]. Electrospinningis a process in which a charged polymer jet is collectedon a grounded collector; a rapidly rotating collector resultsin aligned nanofibers while stationary collectors result inrandomly oriented fiber mats. Electrospinning forms superfinefibers with diameters ranging from 10 µm down to 10 nmby forcing a polymer solution with an electric field through aspinneret [39]. The advantages of electrospun nanofibers overconventional film-casting techniques include a high surfacearea, high mass to volume ratio and a small inter-fibrous poresize with high porosity [40]. These features make electrospunpolymeric fibers excellent candidates for fabrication ofimplant surfaces and for controlled periprosthetic drugdelivery [41–45]. Organic solvents or acids are usuallyrequired to dissolve either synthetic (PLGA and PLA, etc)or natural (chitosan, gelatin, collagen, etc) polymers [46–50]prior to electrospinning [51]. PVA is water-soluble and hasbetter nanofiber forming capability [50, 52], in part dueto its sufficiency in electroconductivity [52, 53]. However,the limitation of PVA nanofibers are obvious, including fasthydrolysis [54] and its bioinert nature [55] that hinder proteinand cell adhesion [56]. Thus, additional efforts are requiredto incorporate nano-HA particles and collagen into PVAnanofibers to mimic natural extracellular bone matrix.
This paper reports a biomimetic nanofiber scaffoldconsisting of PVA–Col with unidirectionally aligned nano-HA particles. The physiochemical properties and degradationprofiles are described. In addition, the behavior of murineosteoblastic MC3T3 cells (adhesion, proliferation anddifferentiation) growing on the PVA-Col-HA nanofiberscaffolds was examined.
2. Materials and methods
2.1. Materials
Nano-HA (average diameter 20–70 nm, specific surfacearea 110 m2 g−1, density 3.0 g ml−1; Berkeley AdvancedBiomaterials, Inc., Berkeley, CA, USA), PVA (Mw ∼ 205 000,86.7–88.7 mol% hydrolysis, 4200 ∼ polymerization), al-bumin fluorescein isothiocyanate conjugate bovine (albu-min FITC; >7% mol FITC albumin), Type I collagen(from rat tail, in 5 wt% acetic acid; Sigma-Aldrich, StLouis, MO, USA), ultrapure water (DNAse, RNAse free,0.1 µm filtered; Invitrogen Corp., Grand Island, NY, USA),murine MC3T3-E1 pre-osteoblasts (ATCC, Manassas, VA,USA), α-modified minimum essential medium (α-MEM;Invitrogen, Carlsbad, CA), USA, alkaline phosphatase(ALP) assay kit (ATCC) and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay kit (ATCC).
2.2. Electrospinning of PVA-Col-HA nanofibers
PVA was dissolved in ultrapure water at 90 ◦C (10%, w/v,g ml−1). The emulsified solution of HA nanoparticles wasadded to PVA solution and homogenized at 60 ◦C in variousHA/PVA ratios (HA/PVA = 1/2, 1/4, 1/9, v/v). AlbuminFITC was firstly dissolved in ultrapure water (0.1 mg ml−1)and then thoroughly mixed with cooled HA–PVA slurry(albumin, 10%, v/v) for 10 min. Collagen solution (aceticacid, 5 wt%) was then homogenized with the mixture tofinalize the PVA-Col-HA-albumin FITC blend (Col, 10%,v/v). The suspension of the blend was immediately loadedin a syringe (5 ml, B-D Scientific, Franklin Lakes, NJ,USA) connected to a high purity tubing (IDEX Health andScience, Oak Harbor, WA, USA) with a 26 G1/2 needle(B-D Scientific) with a blunted tip (0.6 mm inner diameter).The syringe was attached to a syringe pump (R-100E,Razel Scientific Instruments, St Albans, VT, USA) with asetting flow rate Q. A high voltage power supply (ES40P,Gamma High Voltage Research, Inc., Ormond Beach, FL,USA) was connected to the needle. The electrospinningsetting parameters were as follows: voltage = 18 kV, Q =7.8 µl min−1, needle tip-to-collector distance = 10 cm. Arotational speed of 500 rpm was used to evenly depositthe solute on glass coverslips attached to the stainless steelhexahedron collector linked to the shaft of a motorized stirrer(JR4000, Arrow Engineering Co, Inc., Hillside, NJ, USA).Electrospinning was running in the dark for 1 h to collectfibers. Samples were dried overnight (O/N) before testing.
2.3. Characterizations of PVA-Col-HA nanofibers
2.3.1. Scanning electron microscope (SEM). The PVA-Col-HA nanofibers deposited on coverslips were gold-coated(Gold Sputter, EFFA Coater, Redding, CA, USA) and themorphology of the nanofibers was characterized by Scanningelectron microscope (SEM) (JSM-6510LV-LGS, MA, USA).The distribution of fiber diameter was measured from theSEM images using analysis software (Image J, NationalInstitute of Health, Bethesda, MD, USA).
2
Nanotechnology 23 (2012) 115101 W Song et al
2.3.2. Transmission electron microscope (TEM). Theencapsulation of HA nanoparticles in individual fibers wasdetermined with a transmission electron microscope (TEM)(JEOL-2010 FasTEM, Peabody, MA, USA). TEM sampleswere prepared by directly depositing nanofibers onto Cugrids covered with ultrathin carbon layers. Dispersive x-rayspectrometry (EDAX) was used to analyze the chemicalcomposition in the area of the sample surface of interest.
2.3.3. Optical microscope. Images of nanofibers were alsorecorded under 10× and 40× magnification through bothvisible light and fluorescence by an optical microscope (HAL100, Carl Zeiss Microimaging, Inc., Thornwood, NY, USA),supplemented with microscope light source system (X-CiteSeries 120Q, Lumin Dynamics Group Inc., Mississauga,Canada).
The intensity of fluorescence from albumin FITCincorporated into the nanofibers was also quantified byreading the deposited coverslips with the excitation/emissionwavelength of 485/528 nm in a UV/vis spectrophotometer(BioTek Synergy HT, Winooski, VT, USA). Total 9×9 countswere measured to calculate the mean relative fluorescence unit(RFU).
2.3.4. Contact angle. The sessile drop method was usedto measure the contact angle of the coverslip surfaces withdeposited nanofibers at ambient temperature. Three 10 µldroplets of distilled water were placed on the parts ofthe coverslip surface until equilibrium and contact angleswere measured by a contact angle goniometer (Rame-Hart,Mountain Lakes, NJ, USA) to calculate the mean value.
2.3.5. Atomic force microscopy (AFM). We characterizedthe surface morphology of nanofibers using a MultimodeIIIa AFM (Digital Instruments) and a Dimension 3100AFM (VEECO, Santa Barbara, CA, USA). AFM imagingof the nanofibers was conducted using contact mode in air.Positioning of the cantilever on top of the nanofibers withinmicrometer accuracy is achieved by the integrated opticalmicroscope. The data were collected by mapping the fiberswithin a 20× 20 µm2 sized grid.
The nanomechanical properties of the nanofibers canbe studied by AFM force–distance curves in contact forcecalibration mode [57, 58]. The calculation was done byanalyzing data collected on more than 20 spots using sixspecimens. The force curves were fitted to the Hertz modeland Hooke’s law was used to calculate the loading force, F.The nanofibers have a much larger radius of curvature thanthe AFM tip (∼20 nm) and can be treated as a planar surface.The measured elastic modulus, E, is qualitatively related toYoung’s modulus.
2.4. Rheology characterization of nanofiber precursorsolutions
An AR2000 rheometer (TA Instruments Inc., New Castle, DL,USA) with a standard steel parallel-plate geometry of 20 mmdiameter was used for the rheological characterization of all
precursor solutions. Test methods of oscillatory stress sweepand frequency sweep were employed. The stress sweep washolding the temperature (25 ◦C) and frequency (2π rad s−1)constant while increasing the stress level from 0.001 to 10 Pa.The linear viscoelastic region (LVR) from 0.001 to 10 Pawas determined as a safe region without structural breakagefrom oscillatory stress. Samples were subjected to a steadystress ramp and the corresponding storage modulus (G′) andloss modulus (G′′) were measured. The frequency sweep wasperformed at a fixed stress corresponding to the point in themiddle of the LVR profile. The oscillatory frequency wasincreased from 0.1 to 100 rad s−1, and the plots of G′ andG′′ versus frequency were obtained from the manufacturer’ssoftware. Complex viscosity (η∗) and the tangent of thephase angle (tan δ) (ratio of G′′/G′) were obtained from theoscillatory test to plot against frequency as well.
2.5. Degradation of PVA-Col-HA nanofibers
The PVA-Col-HA nanofibers deposited on coverslips weresubjected to in vitro degradation study at 37 ◦C. Briefly, thecoverslips were placed into 12-well plate and immersed inultrapure water in an incubator for 236 h. The morphology ofnanofibers after water-immersion was characterized by opticalmicroscopy and SEM.
2.6. Cell culture
MC3T3 cells were seeded on the coverslips deposited withelectrospun fibers. Briefly, coverslips were placed into 12-wellcell-culturing plates. Murine MC3T3-E1 pre-osteoblasts(ATCC) were cultured in α-MEM (Invitrogen) supplementedwith 10% fetal bovine serum (Invitrogen), 10 mM β-glycerophosphate (Sigma), and a 1% (v/v) antibiotic mixtureof penicillin and streptomycin at 37 ◦C in a humidifiedincubator with 5% CO2. MC3T3-E1 cells were plated at adensity of 1 × 104 cells/well (12-well plate) onto placedcoverslip surfaces with deposited nanofibers.
2.7. Cell attachment and proliferation
The mitochondrial activity of MC3T3 cells cultured withnanofibers after 48 and 72 h was determined by colorimetricassay which detected the conversion of MTT (ATCC) toformazan. MTT solution (1 mg ml−1 in test medium) wasadded (20 µl/well) and the wells were incubated at 37 ◦Cfor 4 h to allow the formation of formazan crystals. Thendimethylsulfoxide (DMSO) (100 µl/well) was added to allwells and mixed thoroughly to dissolve the dark blue crystals.The optical density (OD) of the extracted medium wasmeasured in a UV/vis spectrophotometer (BioTek SynergyHT, USA) at 490 nm.
viability/cytotoxicity staining kit (Invitrogen). Viable cellsfluoresce green through the reaction of calcein AM withintracellular esterase, whereas non-viable cells fluorescered due to the diffusion of ethidium homodimer acrossdamaged cell membranes and binding with nucleic acids. Thecell–nanofiber interaction was visualized under a fluorescence
3
Nanotechnology 23 (2012) 115101 W Song et al
Figure 1. Light-visible images of PVA-Col-HA nanofibers with a HA/PVA ratio of (a), (b) 1/2, (c), (d) 1/4 and (e), (f) 1/9 (v/v).Fluorescence images of the same nanofibers when excited with blue light. Images were observed under (a), (c), (e) 10× and (b), (d), (f) 40×objectives, respectively.
microscope (Carl Zeiss Microimaging) and AFM (DigitalInstruments Nanoscope IIIA).
For the cell differentiation experiment, MC3T3 cells werecultured in osteogenic media (regular media described aboveplus 10 mM β-glycerophosphate and 50 µg ml−1 L-ascorbicacid (Sigma)) before they were harvested at 9 days. Culturemedia were changed every 3 days. The ALP activity ofMC3T3 cells after treating with differentiation medium wasmeasured by a commercial kit (ATCC). Then 80 µl of ALPbuffer was added and the plate was incubated at 4 ◦C for4 h. The cell lysate was then centrifuged at a speed of13 000g for 3 min and cell debris was discarded. 50 µl ofp-nitrophenyl phosphate (pNPP) was added to react with celllysate for 60 min in the dark at room temperature. 20 µl ofstop solution was then added to terminate the ALP activity.The standard curve was drawn by using pNPP of 0, 4, 8, 12,16, 20 nmol/well to react with 10 µl of ALP enzyme solution.The OD was measured at 405 nm.
2.8. Statistical analysis
Data were analyzed with SPSS Version 12.0 (SPSS, Chicago,IL). All values are expressed as mean plus or minus standarddeviation. Analysis of variance (ANOVA) was used to analyzethe experimental data from all the experiments. Statisticalsignificance was set to p < 0.05.
3. Results
3.1. Optical microscopy of PVA-Col-HA nanofibers
In this study, an emulsion of HA nanoparticles washomogenized with PVA solution (10%, w/v) at various ratiosof 1/2, 1/4 and 1/9 (v/v). Randomly oriented electrospunfibers are visualized in figure 1. Micro-beads can be observed
from HA/PVA (1/2) nanofibers, as shown in figure 1(a),which were caused by the failure of fiber formation duringspinning. Apparently, an excessive amount (HA/PVA = 1/2)of incorporated nano-HA disturbed the homogeneity of PVAsolution and the stability of electrospinning. In comparison,fiber mesh fabricated from PVA-Col-HA precursor solutionwith a smaller amount of nano-HA (1/4 and 1/9) was thickerand more homogeneous, as shown in figures 1(c) and (e). Thepresence of albumin FITC coupling with HA nanoparticleswas visualized in electrospun fibers [59, 60].
3.2. Electron microscopy of PVA-Col-HA nanofibers
The encapsulation of nano-HA in nanofibers was character-ized by TEM (figure 2). The rod-like HA particles weremostly oriented along the axial direction inside individualfibers, whereas a few were perpendicular to the axial directionextruding the fiber. The rough surface caused by theseextrusions could provide anchorage sites for cell adhesionduring tissue regeneration [55]. The elemental composition ofnanofibers determined by EDAX reveals the Ca-P componentsencapsulated within fibers. The morphology of fiber structurewas also characterized by SEM (figures 3 and 4). A largeamount of Ca was quantified at the extrusion spots on thenanofiber surfaces.
It is shown that precursor solutions at all HA/PVA ratios(1/2, 1/4 and 1/9) were capable of fabricating fibrous meshesthrough electrospinning (figures 3 and 4). The inclusion ofcollagen during electrospinning showed no effects on fiberformation. The average diameters of fibers without collagenwere similar (508, 484 and 514 nm) at different HA/PVAratios (figure 3), while the average diameters of fibers withcollagen dropped (552, 416 and 362 nm) followed by decreaseof the HA/PVA ratio (figure 4).
4
Nanotechnology 23 (2012) 115101 W Song et al
Figure 2. Representative TEM micrograph of PVA-Col-HA nanofibers with a HA/PVA ratio of 1/4 (upper panels). Elemental compositionof nanofibers obtained from EDAX (bottom panel).
Figure 3. SEM micrograph of PVA-HA nanofibers with a HA/PVA ratio of 1/2, 1/4 and 1/9 (upper panel 1000×; middle panel 10 000×).The diameter histogram of corresponding nanofibers analyzed from 1000× micrographs (bottom panel).
5
Nanotechnology 23 (2012) 115101 W Song et al
Figure 4. SEM micrograph of PVA-Col-HA nanofibers with a HA/PVA ratio of 1/2, 1/4 and 1/9 (upper panel 1000×; middle panel10 000×). The diameter histogram of corresponding nanofibers analyzed from 1000× micrographs (bottom panel).
3.3. Atomic force microscopy (AFM) of PVA-Col-HAnanofibers
We used AFM to study the morphology and mechanicalproperty of single strand fibers with different ratios ofHA/PVA via the contact force model. The tip is kept close tothe sample surface by a feedback mechanism to return heightand deflection images from the scanning area. The roughsurface of the nanofibers was observed from height imagesas expected from the extrusion of encapsulated HA nanorods(figures 5(A) and (B)). The fiber surface became smootherconsistent with reduced encapsulation of HA.
The obtained Young’s modulus (E) of nanofibers withencapsulated HA was higher than that of the pure PVAnanofibers (0.826 MPa), which revealed the incrementof the fiber stiffness. E significantly increased when agreater amount of HA was encapsulated into the nanofibers(HA/PVA = 1/2) (figure 5(C)) compared with other ratios(1/4 and 1/9). The incorporation of Col showed less influenceon the stiffness of the nanofibers.
3.4. Hydrolytic degradation of PVA-Col-HA nanofibers
In this study, all PVA-Col-HA nanofibers were preparedunder the same conditions with various ratios of HA/PVA.PVA shows high efficiency in electrospinning due to itsstrong hydrophilicity but a high dissolution rate in aqueous
solution [54]. We attempted to incorporate HA nanoparticlesto reinforce the mechanical strength and hydrophobicity ofPVA nanofibers. Nanofibers with different HA/PVA ratios(1/2, 1/4 and 1/9) were immersed in ultrapure water andobserved with a fluorescence microscope (figure 6) and SEM(figure 7(A)). Within 2 h of soaking, the fluorescence intensityof 1/9 nanofibers was low, while others were still intense(figure 6). The fluorescence of nanofibers gradually decayedfollowing long-term soaking (48 and 236 h). The fluorescenceof 1/4 nanofibers was most durable, and the mesh structure of1/4 nanofibers was still observable after soaking for 236 h.
The morphological change of individual fibers aftersoaking can be visualized in SEM micrographs (figure 7(A)).After 48 h of soaking, the fiber structure of 1/2 and 1/4nanofibers retained (figure 7(A)) while 1/9 nanofibers hadmostly failed. Precipitates observed around fibers were HAnanoparticles released from nanofibers during degradation.After 236 h of soaking, the polymeric components of fiberswere mostly degraded, and the encapsulated HA nanoparticlesaligned along the fiber axis.
3.5. Contact angle of PVA-Col-HA nanofibers
The nano-scale roughness on the substrate surface could alterthe hydrophilicity of materials. In this study, encapsulated HAnanoparticles were expected to influence the hydrophilicity
6
Nanotechnology 23 (2012) 115101 W Song et al
Figure 5. AFM contact mode images of (A) PVA-HA and (B) PVA-Col-HA nanofibers with HA/PVA ratios of 1/2, 1/4 and 1/9 (upperpanel, height images; middle panel, deflection images). (C) The mean Young’s modulus (E) of single nanofibers (n = 6). ∗p < 0.05 relatedto 1/2 vs. 1/4 and 1/9 fibers with Col.
of the PVA fiber mesh. The contact angles (θ ) for thesenanofiber meshes (1/2, 1/4 and 1/9) were measured to revealthe hydrophilicity (figure 7(B)). Generally, the contact angledecreases when the amount of encapsulated HA is smaller
for nanofibers with or without collagen. Fibers with collagenhave larger θ than fibers without collagen, probably due toits peptide strands. This result indicated that encapsulated HAnanorods enhanced the hydrophobicity of PVA nanofibers.
7
Nanotechnology 23 (2012) 115101 W Song et al
Figure 6. The fluorescence images (10×) of PVA-Col-HA nanofibers with HA/PVA ratios of 1/2, 1/4 and 1/9 soaking in water at 37 ◦Cfor 2 h (upper panel), 48 h (middle panel) and 236 h (bottom panel).
3.6. Rheological properties of PVA-Col-HA nanofiberprecursor solution
The predominance of G′′ over G′ in all precursor solutions canbe observed from the oscillation stress sweep for the appliednon-destructive stress range (figure 8(a)), which indicated theprevalence of viscous fluid behaviors for precursor solutionin a sol state. The gradual decrease of G′′ followed bythe increase in the amount of HA incorporated showedthe influence of HA on the rheological properties of theprecursor solutions (figure 8(a)). The Col content seemed tominimize G′′ that might indicate its interaction with othermolecules (figure 8(a)). An increase in both G′ and G′′ asthe frequency increased was revealed (figures 8(c) and (d)).All G′′ (1–100 Pa) are dominant over G′ (0.1–10 Pa) andthe decrease in the slope of G′ according to the increasein incorporated HA revealed the weakly structured system.The complex viscosities (η∗) of all solutions were decreasingacross the frequency range, suggesting pseudoplastic behavior(figure 8(b)). It is noted that the slope of η∗ curve acrossthe frequency range was flattened followed by the decreaseof incorporated HA in the system (except 1/2 No-Col). Thisresult suggested more bonding interactions were formed bymore incorporated HA with PVA, and the dismantling ofHA–PVA linkages by higher frequencies of shear stress wasmore significantly revealed by the decrease of η∗. As tan δ isa ratio of G′′ to G′, a value above 1 indicates a material withdominant liquid-like behavior and a value below 1 indicates
a material with dominant elastic (solid-like) behavior [61].All curves of tan δ were above 1, indicating the liquid stateof the precursor solution. However, the addition of nano-HAseemed to assist the transition from sol to gel, as the tan δcurve of the solution with higher HA content gradually felland closely approached 1 at lower frequencies (except 1/2No-Col) (figure 8(c)). Taken together, the incorporated HAwas capable of modifying the viscoelasticity of the precursorsolution by interacting with PVA.
3.7. Cell behavior when cultured on PVA-Col-HA nanofibers
Live–dead cell staining was utilized to study the cellmorphology and interaction with nanofibers after culturing for72 h (figures 9(a) and (b)). The attachment and proliferation oflive MC3T3 cells (green) seeding on the nanofiber mesh canbe visualized, showing good compatibility of cells and fibers.Cells grown on both 1/2 and 1/4 nanofibers (figures 9(a)and (b)) had much better adhesion and cell spreading patternthan the cells grown on 1/9 nanofibers after 72 h of culturing.No obvious dead cells (red) were observed in any of the cellcolonies seeding on nanofibers with and without collagen.Nevertheless, cells on 1/9 nanofibers were less spread, whichis probably due to the earlier failure of the supporting fibermesh.
The MTT result from 8 days of culturing was quiteconsistent with the live–dead staining (figure 9(c)). Cellsseeded on all groups of nanofibers showed a significantly
8
Nanotechnology 23 (2012) 115101 W Song et al
Figure 7. (A) SEM micrograph (10 000×) of PVA-Col-HA nanofibers with a HA/PVA ratio of 1/2 (left) and 1/4 (right) soaking in waterat 37 ◦C for 48 h (upper panel) and 236 h (bottom panel). (B) Contact angles of PVA-HA (red) and PVA-Col-HA (black) nanofibers withHA/PVA ratios of 1/2, 1/4 and 1/9. (Control: glass coverslips without nanofibers.) (C) Representative images of water droplets wetting thesurface of the nanofiber mesh.
higher proliferation rate than cells seeded on coverslipswithout nanofibers (Control). This indicated that thenano-fiber mesh promoted the spread and proliferation ofosteoblastic cells. The incorporation of Col can further benefitcell proliferation along the nanofibers, as shown as thehigher proliferation rate in the MTT results (figure 9(c)).The sudden decrease of cell proliferation on 1/9 nanofiberwith Col might be attributed to cell detachment fromfailed PVA–Col fibers. The 1/4 nanofiber/Col ratio wasrevealed to be most durable and interconnective for celladhesion and proliferation, verified by both live–dead imagesand MTT results (figures 9(a)–(c)). Moreover, the ALPlevels normalized with protein levels were measured foreach cell group to analyze the influence of nanofibers oncell functionality (figure 9(d)). The results showed thatthere was no significant difference among cells culturedon different nanofibers. In sum, the nanofibers promoteadhesion and proliferation of osteoblastic cells withoutinfluencing the cells’ functionality. The cell proliferationrate on nanofibers was dependent on the durability ofthe nanofibers. The cell–nanofiber interaction was further
characterized by contact mode AFM (figure 10). Cellsinteracted with surrounding fibers and infiltrated the fibrousmesh to form a sandwich mode.
4. Discussion
Failure of osseointegration (direct anchorage of an implant bybone formation at the bone–implant surface) represents oneof the main reasons for implant failure and loosening [62].One of the important challenges in the field of implantsurface fabrication is the development of ‘bone-like’ implantsurface substitutes that are ‘intelligent’ and can instructthe in vivo environment to form bone. We propose thata ‘bone-like’ nanofiber implant surface coating mimics thebiological, structural and mechanical behavior of naturalbone, and enhances the adhesion, growth and differentiationof the surrounding bone marrow stromal cells. To imitatethe architecture of the natural bone matrix, we developedelectrospun PVA nanofibers with embedded nano-scale HAand Col. We found that 15% PVA mixed with HA (10%)and Col (10%) provides a better cellular response (adhesion,
9
Nanotechnology 23 (2012) 115101 W Song et al
Figure 8. (a) Storage modulus (G′) and loss modulus (G′′) profiles across an oscillatory range of 0.001–10 Pa at 25 ◦C. (b) Complexviscosity, (c) storage modulus (G′) and (d) loss modulus (G′′) profiles. (e) tan delta across a frequency range of 0.1–10 rad s−1 at 25 ◦C forPVA-HA and PVA-Col-HA precursor solution (HA/PVA = 1/2, 1/4 and 1/9).
proliferation and differentiation), defined degradation pattern(figures 6 and 7), and favorable surface roughness as wellas mechanical strength (figures 3–5) with characterizedphysiochemical properties (figures 5 and 8). PVA-Col-HAnanofibers were prepared using empirical combinationsof electrospinning variables. Several distinct morphologieswere observed, including fiber diameter distribution, surfaceroughness, alignment, orientation of nano-HAs in the fibers,etc (figures 1–5).
PVA has recently been used in electrospun nanofibersfor its hydrophilicity and high electroconductivity, ensuringits feasibility of spinning. Nevertheless, the rapid hydrolysisrate limits the application of PVA electrospun nanofibers asstable and reliable tissue scaffolds. There have been numerousattempts to modify the hydrophilicity of PVA nanofibers. Liet al [54] stabilized PVA nanofibers in water by methanoltreatment to increase the crystallinity of PVA. Pisuchpenet al [63] enhanced the rigidity and hydrophobicity of PVAnanofibers via the addition of silica. Feng and Jiang [64]
showed that highly aligned PVA nanofibers possessedsuperhydrophobicity attributed to the reorientation of PVAmolecules. All of these studies indicate that amphiphilicPVA is inherently chemically alterable. The hydrophilicphenomenon of PVA contacting with water results from theexcessive exposure of hydrophilic hydroxyl groups (–OH) towater molecules and hence hydrogen bonding interactions.Hydrophobic alkyl groups (–CH–), as the backbone of PVA,are repelled by H2O molecules to the interior [65]. The initialswelling stage of PVA in water can be considered as theprocess of conformational reorientation of PVA molecules.While the later dissolving stage represents the stabilization ofreoriented PVA molecules in the water phase. In native PVA,–CH– and –OH groups are evenly distributed around the PVAsurface. However, in certain circumstances, if the arrangementof –CH– groups became dominant on the PVA surface, thelower free surface energy could dramatically alter its originalhydrophilicity in water [64].
10
Nanotechnology 23 (2012) 115101 W Song et al
Figure 9. MC3T3 cell activity assay. The live (green)–dead (red) staining ((a), (b) objectives, 10× and 40×, respectively) of MC3T3 cellscultured on PVA-HA (a) and PVA-Col-HA (b) nanofiber meshes for 72 h. MTT results (c) of cells cultured on HA-PVA-albumin FITC(white column) and PVA-Col-HA (black column) nanofibers for 8 days. Normalized ALP activity of cells cultured on PVA-HA (whitecolumn) and PVA-Col-HA (black column) nanofibers for 8 days. Control: glass coverslips without nanofibers. ∗∗ p < 0.005, n = 3.
It has been demonstrated the encapsulated nano-HA andCol in PVA nanofibers would form hydrogen bonds withPVA molecules, which significantly changed the crystallinityof PVA [55]. Hydrogen bonding in the blended system can
also be demonstrated from the rheological characterization(figure 8). Our results indicated that the PVA-Col-HAprecursor solutions were in a viscous sol-like state that iscapable of being electrospun. The interaction of embedded
11
Nanotechnology 23 (2012) 115101 W Song et al
Figure 10. AFM contact mode images of MC3T3 cells after interaction with PVA-Col-HA nanofibers with a HA/PVA ratio of 1/2 (leftpanel) and 1/4 (right panel) after culturing for 8 days.
nano-HA and Col with PVA might result in the reductionof free –OH groups while augmenting –CH– groups on thenanofiber surface, subsequently affecting the wettability ofthe nanofibers. Our study has demonstrated that the waterrepelling effect of nanofibers attributed to incorporation ofnano-HA was increased with changing the HA ratio (fromHA/PVA = 1/9 to 1/2), as shown in contact angle results(figure 7(B)). More gradual change and larger contact anglescould be revealed from nanofibers consisting of Col, whichalso confirmed the hydrogen bonding interaction of Col–PVA.From aspects of nanostructured surface-interface theory (theCassie–Baxter model), a rough surface allows air to be trappedbetween the liquid and the underlying solid surface, repellingthe water [63]. In the present study, the encapsulation of HAnanorods within the nanofibers was identified to increase thesurface roughness (figures 3–5). The contact angle of theHA–PVA nanofiber surface was probably increased due to thehydrophobic pockets located on the interface.
The promotion of hydrolytic resistance enabled thesupportive role of nanofibers, allowing cell infiltration andmigration. Studies have shown that cellular growth andfunctions can be regulated by nanofiber parameters, such asalignment, pore geometry, surface area, etc. In physiologicalconditions, ECM fibers provide cells with topographicalfeatures that trigger morphogenesis by interaction with thecells through their transmembrane integrin receptors to initiateintracellular signaling cascades [66, 67]. Nanofibers preparedas bone scaffolds were revealed to interact with bone cells andguide mineralization [47]. We propose that the encapsulated
nano-HA and Col provide anchorage sites for cell adhesionand permit nanofibers to mechanically support cells. Thecell–nanofiber interaction can be visualized by cell stainingand AFM images (figures 9 and 10)
One of the potential applications of this nanofiberscaffold is to provide a ‘bone-like’ implant surface coatingto enhance osseointegration by allowing the recruitment ofbone marrow osteoblastic cells into the scaffold matrix.However, there are some limitations need to be addressed forthis study. The efficiency of electrospinning was estimatedto be interrupted by adding HA nanoparticles and Col tothe PVA precursor solution. Parameters of the precursorsolution for electrospinning, such as viscosity, molecularweight, electroconductivity, etc, could dramatically influencethe generation of nanofibers. Further investigations on somedecisive parameters corresponding to the HA/Col ratio needto be performed in the future. Moreover, the incorporationof HA and Col extended the degradation of PVA nanofibersto about 10 days. But more durable fibers applied asbiomimetic tissue engineering scaffolds could be furtherstabilized by cross-linking approaches, such as UV treatment,freeze–thawing, etc, and exploration of tunable degradabilityof nanofibers need to be elaborated in the future.
5. Conclusion
PVA-Col-HA nanofibers fabricated by electrospinningrevealed excellent biodegradability, biocompatibility andbioactivity. The mechanical strength and resistance to
12
Nanotechnology 23 (2012) 115101 W Song et al
hydrolysis of the nanofibers are reinforced by incorporationof HA nanoparticles and collagen. The nanofibers enhancedthe adhesion and proliferation of interacting bone cells. Wepropose that PVA-Col-HA nanofibers might be promisingbiomimetic materials modifying the implant surface fororthopedic applications.
Acknowledgments
This research was supported by Detroit Medical Cen-ter/Providence Orthopedic Residency Program ResearchFunds. We would like to acknowledge Dr Mao’s stu-dent Yi Zou for sample characterization, and the GrantsDMR-0216084 and CTS-0216109 from the National ScienceFoundation.
References
[1] Berry D J, Harmsen W S, Cabanela M E and Morrey B F 2002Twenty-five-year survivorship of two thousand consecutiveprimary charnley total hip replacements: factors affectingsurvivorship of acetabular and femoral components BoneJoint Surg. Am. 84 171–7
[2] Garellick G, Malchau H and Herberts P 2000 Survival of hipreplacements: a comparison of a randomized trial and aregistry Clin. Orthop. 375 157–67
[3] Engh C A, Massin P and Suthers K E 1990 Roentgenographicassessment of the biologic fixation of porous-surfacedfemoral components Clin. Orthop. 257 107–28
[4] Kadoya Y, Revell P A, Al Saffar N, Kobayashi A, Scott G andFreeman M A 1996 Bone formation and bone resorption infailed total joint arthroplasties: histomorphometric analysiswith histochemical and immunohistochemical techniqueJ. Orthop. Res. 14 473–82
[5] Scott D F and Jaffe W L 1996 Host-bone response toporous-coated cobalt-chrome and hydroxyapatite-coatedtitanium femoral components in hip arthroplasty.Dual-energy x-ray absorptiometry analysis of pairedbilateral cases at 5 to 7 years J. Arthroplasty 11 429–37
[6] Stanishevsky A, Chowdhury S, Chinoda P and Thomas V 2008Hydroxyapatite nanoparticle loaded collagen fibercomposites: microarchitecture and nanoindentation studyJ. Biomed. Mater. Res. A 86A 873–82
[7] Wei G B and Ma P X 2004 Structure and properties ofnano-hydroxyapatite/polymer composite scaffolds for bonetissue engineering Biomaterials 25 4749–57
[8] Pramanik N, Mishra D, Banerjee I, Maiti T K, Bhargava P andPramanik P 2009 Chemical synthesis, characterization, andbiocompatibility study of hydroxyapatite/chitosanphosphate nanocomposite for bone tissue engineeringapplications Int. J. Biomater. 2009 512417
[9] Park J W 2011 Increased bone apposition on a titanium oxidesurface incorporating phosphate and strontium Clin. OralImplants Res. 22 230–4
[10] Gittens R A, McLachlan T, Olivares-Navarrete R, Cai Y,Berner S, Tannenbaum R, Schwartz Z, Sandhage K H andBoyan B D 2011 The effects of combinedmicron-/submicron-scale surface roughness and nanoscalefeatures on cell proliferation and differentiationBiomaterials 32 3395–403
[11] Hermida J C, Bergula A, Dimaano F, Hawkins M,Colwell C W Jr and D’Lima D D 2010 An in vivoevaluation of bone response to three implant surfaces usinga rabbit intramedullary rod model J. Orthop. Surg. Res. 5 57
[12] Junker R, Dimakis A, Thoneick M and Jansen J A 2009Effects of implant surface coatings and composition on
bone integration: a systematic review Clin. Oral ImplantsRes. 20 (Suppl 4) 185–206
[13] de Jonge L T, Leeuwenburgh S C, Wolke J G andJansen J A 2008 Organic–inorganic surface modificationsfor titanium implant surfaces Pharm. Res. 25 2357–69
[14] Borsari V, Fini M, Giavaresi G, Rimondini L, Consolo U,Chiusoli L, Salito A, Volpert A, Chiesa R andGiardino R 2007 Osteointegration of titanium andhydroxyapatite rough surfaces in healthy and compromisedcortical and trabecular bone: in vivo comparative study onyoung, aged, and estrogen-deficient sheep J. Orthop. Res.25 1250–60
[15] Perla V and Webster T J 2005 Better osteoblast adhesion onnanoparticulate selenium—a promising orthopedic implantmaterial J. Biomed. Mater. Res. A 75 356–64
[16] Omar O, Lenneras M, Svensson S, Suska F, Emanuelsson L,Hall J, Nannmark U and Thomsen P 2010 Integrin andchemokine receptor gene expression in implant-adherentcells during early osseointegration J. Mater. Sci. Mater.Med. 21 969–80
[17] Zhang C, Li Z A, Cheng X R, Xiao Q and Li H B 2010Hydroxyapatite crystals biologically inspired on titanium byusing an organic template based on the copolymer of acrylicacid and itaconic acid J. Biomed. Mater. Res. A 92 63–9
[18] Tan K S, Qian L, Rosado R, Flood P M and Cooper L F 2006The role of titanium surface topography on J774A.1macrophage inflammatory cytokines and nitric oxideproduction Biomaterials 27 5170–7
[19] De R M, Carteni’ M, Petillo O, Calarco A, Margarucci S,Rosso F, De R A, Farina E, Grippo P and Peluso G 2004Cationic polyelectrolyte hydrogel fosters fibroblastspreading, proliferation, and extracellular matrixproduction: implications for tissue engineering J. Cell.Physiol. 198 133–43
[20] Kim S S, Park M S, Jeon O, Choi C Y and Kim B S 2006Poly(lactide-co-glycolide)/hydroxyapatite compositescaffolds for bone tissue engineering Biomaterials27 1399–409
[21] Estey T, Kang J, Schwendeman S P and Carpenter J F 2006BSA degradation under acidic conditions: A model forprotein instability during release from PLGA deliverysystems J. Pharmaceut. Sci. 95 1626–39
[22] Fu K, Pack D W, Klibanov A M and Langer R 2000 Visualevidence of acidic environment within degradingpoly(lactic-co-glycolic acid) (PLGA) microspheresPharmaceut. Res. 17 100–6
[23] So K, Takemoto M, Fujibayashi S, Neo M, Kyomoto M,Hayami T, Hyon S H and Nakamura T 2007Antidegenerative effects of partial disc replacement in ananimal surgery model Spine 32 1586–91
[24] Matsumura K, Hayami T, Hyon S H and Tsutsumi S 2009Control of proliferation and differentiation of osteoblasts onapatite-coated poly(vinyl alcohol) hydrogel as an artificialarticular cartilage material J. Biomed. Mater. Res. A92 1225–32
[25] Ngawhirunpat T, Opanasopit P, Rojanarata T,Akkaramongkolporn P, Ruktanonchai U andSupaphol P 2009 Development of meloxicam-loadedelectrospun polyvinyl alcohol mats as a transdermaltherapeutic agent Pharm. Dev. Technol. 14 70–9
[26] Sailaja G S, Sreenivasan K, Yokogawa Y, Kumary T V andVarma H K 2009 Bioinspired mineralization and celladhesion on surface functionalized poly(vinyl alcohol)films Acta Biomater. 5 1647–55
[27] Gittens R A, McLachlan T, Olivares-Navarrete R, Cai Y,Berner S, Tannenbaum R, Schwartz Z, Sandhage K H andBoyan B D 2011 The effects of combinedmicron-/submicron-scale surface roughness and nanoscalefeatures on cell proliferation and differentiationBiomaterials 32 3395–403
[28] Mendonca G, Mendonca D B, Simoes L G, Araujo A L,Leite E R, Duarte W R, Aragao F J and Cooper L F 2009The effects of implant surface nanoscale features onosteoblast-specific gene expression Biomaterials30 4053–62
[29] Dalby M J, McCloy D, Robertson M, Wilkinson C D andOreffo R O 2006 Osteoprogenitor response to definedtopographies with nanoscale depths Biomaterials27 1306–15
[30] Dohan Ehrenfest D M, Coelho P G, Kang B S, Sul Y T andAlbrektsson T 2010 Classification of osseointegratedimplant surfaces: materials, chemistry and topographyTrends Biotechnol. 28 198–206
[31] Tomisa A P, Launey M E, Lee J S, Mankani M H,Wegst U G and Saiz E 2011 Nanotechnology approaches toimprove dental implants Int. J. Oral Maxillofac Implants26 (Suppl) 25–44
[32] Gittens R A, McLachlan T, Olivares-Navarrete R, Cai Y,Berner S, Tannenbaum R, Schwartz Z, Sandhage K H andBoyan B D 2011 The effects of combinedmicron-/submicron-scale surface roughness and nanoscalefeatures on cell proliferation and differentiationBiomaterials 32 3395–403
[33] Dohan Ehrenfest D M, Coelho P G, Kang B S, Sul Y T andAlbrektsson T 2010 Classification of osseointegratedimplant surfaces: materials, chemistry and topographyTrends Biotechnol. 28 198–206
[34] Huang Z, Daniels R H, Enzerink R J, Hardev V, Sahi V andGoodman S B 2008 Effect of nanofiber-coated surfaces onthe proliferation and differentiation of osteoprogenitors invitro Tissue Eng. A 14 1853–9
[35] Prabhakaran M P, Venugopal J and Ramakrishna S 2009Electrospun nanostructured scaffolds for bone tissueengineering Acta Biomater. 5 2884–93
[36] Asran A S, Henning S and Michler G H 2010 Polyvinylalcohol-collagen-hydroxyapatite biocomposite nanofibrousscaffold: mimicking the key features of natural bone at thenanoscale level Polymer 51 868–76
[37] Francis L, Venugopal J, Prabhakaran M P, Thavasi V,Marsano E and Ramakrishna S 2010 Simultaneouselectrospin-electrosprayed biocomposite nanofibrousscaffolds for bone tissue regeneration Acta Biomater.6 4100–9
[38] Schofer M D, Fuchs-Winkelmann S, Grabedunkel C, Wack C,Dersch R, Rudisile M, Wendorff J H, Greiner A,Paletta J R and Boudriot U 2008 Influence of poly(L-lacticacid) nanofibers and BMP-2-containing poly(L-lactic acid)nanofibers on growth and osteogenic differentiation ofhuman mesenchymal stem cells Sci. World J. 8 1269–79
[39] Torres-Giner S, Gimeno-Alcaniz J V, Ocio M J andLagaron J M 2009 Comparative performance of electrospuncollagen nanofibers cross-linked by means of differentmethods ACS Appl. Mater. Interfaces 1 218–23
[40] Sundararaghavan H G, Metter R B and Burdick J A 2010Electrospun fibrous scaffolds with multiscale andphotopatterned porosity Macromol. Biosci. 10 265–70
[41] Kim G M, Asran A S, Michler G H, Simon P andKim J S 2008 Electrospun PVA/HAp nanocompositenanofibers: biomimetics of mineralized hard tissues at alower level of complexity Bioinspir. Biomim. 3 046003
[42] Sargeant T D, Guler M O, Oppenheimer S M, Mata A,Satcher R L, Dunand D C and Stupp S I 2008 Hybrid boneimplants: self-assembly of peptide amphiphile nanofiberswithin porous titanium Biomaterials 29 161–71
[43] Asran A S, Henning S and Michler G H 2010 Polyvinylalcohol-collagen-hydroxyapatite biocomposite nanofibrousscaffold: mimicking the key features of natural bone at thenanoscale level Polymer 51 868–76
[44] Cho W J, Kim J H, Oh S H, Nam H H, Kim J M andLee J H 2009 Hydrophilized polycaprolactone nanofibermesh-embedded poly(glycolic-co-lactic acid) membranefor effective guided bone regeneration J. Biomed. Mater.Res. A 91 400–7
[45] Ekaputra A K, Zhou Y, Cool S M and Hutmacher D W 2009Composite electrospun scaffolds for engineering tubularbone grafts Tissue Eng. A 15 3779–88
[46] Duan B, Wu L L, Li X R, Yuan X Y, Li X L, Zhang Y andYao K D 2007 Degradation of electrospunPLGA-chitosan/PVA membranes and theircytocompatibility in vitro J. Biomater. Sci.-Polym. Edn18 95–115
[47] Yoshimoto H, Shin Y M, Terai H and Vacanti J P 2003 Abiodegradable nanofiber scaffold by electrospinning and itspotential for bone tissue engineering Biomaterials24 2077–82
[48] Huang L, Nagapudi K, Apkarian R P and Chaikof E L 2001Engineered collagen-PEO nanofibers and fabricsJ. Biomater. Sci.-Polym. Edn 12 979–93
[49] Shin Y M, Hohman M M, Brenner M P andRutledge G C 2001 Experimental characterization ofelectrospinning: the electrically forced jet and instabilitiesPolymer 42 9955–67
[50] Zhou Y S, Yang D Z, Chen X M, Xu Q, Lu F M andNie J 2008 Electrospun water-soluble carboxyethylchitosan/poly(vinyl alcohol) nanofibrous membrane aspotential wound dressing for skin regenerationBiomacromolecules 9 349–54
[51] Huang Z M, Zhang Y Z, Kotaki M and Ramakrishna S 2003A review on polymer nanofibers by electrospinning andtheir applications in nanocomposites Compos. Sci. Technol.63 2223–53
[52] Wu L L, Yuan X Y and Sheng J 2005 Immobilization ofcellulase in nanofibrous PVA membranes byelectrospinning J. Membr. Sci. 250 167–73
[53] Kenawy E R, Layman J M, Watkins J R, Bowlin G L,Matthews J A, Simpson D G and Wnek G E 2003Electrospinning of poly(ethylene-co-vinyl alcohol) fibersBiomaterials 24 907–13
[54] Yao L, Haas T W, Guiseppi-Elie A, Bowlin G L,Simpson D G and Wnek G E 2003 Electrospinning andstabilization of fully hydrolyzed poly(vinyl alcohol) fibersChem. Mater. 15 1860–4
[55] Asran A S, Henning S and Michler G H 2010 Polyvinylalcohol-collagen-hydroxyapatite biocomposite nanofibrousscaffold: mimicking the key features of natural bone at thenanoscale level Polymer 51 868–76
[56] Schmedlen K H, Masters K S and West J L 2002Photocrosslinkable polyvinyl alcohol hydrogels that can bemodified with cell adhesion peptides for use in tissueengineering Biomaterials 23 4325–32
[57] Blacklock J, Sievers T K, Handa H, You Y Z, Oupicky D,Mao G and Mohwald H 2010 Cross-linked bioreduciblelayer-by-layer films for increased cell adhesion andtransgene expression J. Phys. Chem. B 114 5283–91
[58] Mermut O, Lefebvre J, Gray D G and Barrett C J 2003Structural and mechanical properties of polyelectrolytemultilayer films studied by AFM Macromolecules36 8819–24
[59] Hlady V and Furedimilhofer H 1979 Adsorption ofhuman-serum albumin on precipitated hydroxyapatiteJ. Colloid Interface Sci. 69 460–8
[60] Wassell D T H, Hall R C and Embery G 1995 Adsorption ofbovine serum-albumin onto hydroxyapatite Biomaterials16 697–702
[61] Tanna S, Taylor M J, Sahota T S and Sawicka K 2006Glucose-responsive UV polymerised dextran-concanavalin
A acrylic derivatised mixtures for closed-loop insulindelivery Biomaterials 27 1586–97
[62] Ramaswamy Y, Wu C and Zreiqat H 2009 Orthopedic coatingmaterials: considerations and applications Expert Rev. Med.Devices 6 423–30
[63] Pisuchpen T, Chaim-Ngoen N, Intasanta N, Supaphol P andHoven V P 2011 Tuning hydrophobicity and water adhesionby electrospinning and silanization Langmuir 27 3654–61
[64] Feng L, Song Y L, Zhai J, Liu B Q, Xu J, Jiang L andZhu D B 2003 Creation of a superhydrophobic surface froman amphiphilic polymer Angew. Chem.-Int. Edn 42 800–2
[65] Hassan C M and Peppas N A 2000 Structure and applicationsof poly(vinyl alcohol) hydrogels produced by conventionalcrosslinking or by freezing/thawing methods Adv. Polym.Sci. 153 37–65
[66] Kim D H, Lipke E A, Kim P, Cheong R, Thompson S,Delannoy M, Suh K Y, Tung L and Levchenko A 2010Nanoscale cues regulate the structure and function ofmacroscopic cardiac tissue constructs Proc. Natl Acad. Sci.USA 107 565–70
[67] Hynes R O 2009 The extracellular matrix: not just prettyfibrils Science 326 1216–9
