Colloids and Surfaces B: Biointerfaces · Colloids and Surfaces B: Biointerfaces 75 (2010) 107–114 Contents lists available at ScienceDirect Colloids and Surfaces B: Biointerfaces

Page 1

I

Ca

b

a

ARRAA

KCMPRS

1

igiattvgtbatTib

T

0d

Colloids and Surfaces B: Biointerfaces 75 (2010) 107–114

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journa l homepage: www.e lsev ier .com/ locate /co lsur fb

mpact of RGD micro-patterns on cell adhesion

. Cholleta,∗, S. Lazareb, F. Guillemota, M.C. Durrieua

INSERM, U577, Bordeaux F-33076, Univ. Victor Segalen Bordeaux 2, 146 rue Léo Saignat, 33076 Bordeaux Cedex, FranceInstitut des Sciences Moléculaires (ISM), UMR 5255 du CNRS, Université de Bordeaux 1, 351 cours de la Libération, F-33405 Talence, France

r t i c l e i n f o

rticle history:eceived 2 December 2008eceived in revised form 6 April 2009ccepted 12 August 2009vailable online 20 August 2009

eywords:ell adhesionicro-patterning

oly-(ethylene terephthalate)

a b s t r a c t

In order to avoid the problems related to biomaterial use (inflammation, infections, aseptic loosening,etc.), a new approach consisting of associating the material and autologous cells before implantationis being developed, thus requiring a perfect cooperation between the material’s surface and the cell.To improve cell adhesion to biomaterials, a suitable method is to functionalize their surface by pro-adhesive ligand grafting. The aim of this study was to covalently graft RGD containing peptides ontoa poly-(ethylene terephthalate) surface in well-defined microstructures in order to control MC3T3 celladhesion. We followed two different routes for obtaining micro-patterned materials: (1) a photoabla-tion technique using an excimer laser and (2) a photolithography process. The resulting patterns werecharacterized by optical microscopy, scanning electron microscopy, optical profilometry and high reso-

GD peptidesurface modification

lution �-imager. The biological evaluation of cell adhesion onto the micro-patterned surfaces was carriedout using optical microscopy, scanning electron microscopy and fluorescence microscopy. Cells seededonto photolithographical or photoablated micro-patterned PET exhibited an alignment with the RGDdomains and appear to be connecting through pseudopods extending towards each other. Whateverthe technique used to create micro-patterns, a cell alignment occurs once the thickness of the RGD linereaches ∼100 �m. These results prove the importance of microstructured surfaces for the elaboration oftissue engineered biomaterials.

© 2009 Published by Elsevier B.V.

. Introduction

Polyethylene terephthalate (PET) is widely used for cell cultur-ng (as Thermanox® or Mylar®), surgical suture material, vascularrafts (as Dacron®) and anterior cruciate ligament prosthesis due tots biocompatibility and its desirable mechanical properties, suchs strength and resistance. However, this material presents, in longerm applications, a random success rate due to their lack of bioin-egration in the surrounding tissues. For example, the use of PET inascular surgery presents a patency rate of 93% as aortic bifurcationrafts, but its success rate declines drastically for small diame-er vessels. Moreover, PET is not a totally inert material and cane susceptible to chemical attack and hydrolysis that may causelterations in its mechanical properties. However, these proper-

ies can be improved by the means of surface modification [1,2].hus, over the last few years, researchers have focused their stud-es on biomaterial surface modifications in order to improve theiriocompatibility [3,4]. A recent method is to associate a cell popu-

∗ Corresponding author at: INSERM, U577, Bordeaux F-33076, France.el.: +33 557571730; fax: +33 556900517.

E-mail address: [email protected] (C. Chollet).

927-7765/$ – see front matter © 2009 Published by Elsevier B.V.oi:10.1016/j.colsurfb.2009.08.024

lation to these biomaterials before implantation in order to obtainperfect biocompatibility [5,6]. But this “hybrid material” conceptrequires that the surface becomes more “attractive” for the cellsthat will colonise it. To do this, one suitable approach is to function-alize the surface of the materials by “pro-adhesive” ligands likelyto improve cell adhesion and hence the cell and tissue colonisa-tion of these materials. The cell responses are then controlled byintra-cellular signalling pathways [7] that are originally triggeredby transmembrane proteins interacting with the engineered sur-face [8]. Since 1984, RGD peptides have been known to enhancecell attachment on biomaterial surfaces. This attachment propertyis due to the presence of integrin receptors implied in the cel-lular adhesion phenomena [9]. Today, a number of studies haveproven that the immobilization of adhesive peptides on materi-als enhances endothelial [10] and osteoblastic [11] cell adhesion.However, the surface chemistry of such engineered materials, char-acterized by the type of cell-binding ligands (peptides, proteins,etc.), their surface density [12,13] and spatial distribution as well as

their conformation has been demonstrated to be important surfacecues. Recently, Maheshwari et al. demonstrated that cell motil-ity may be regulated by varying ligand spatial presentation at thenanoscale level, and suggested that integrin clustering is requiredto support cell locomotion [14].

Page 2

1 ces B:

iIcpttcsmma2a[tooaoamn[hfsta

hhfew

au(Mlrpm(sf[bttfCccgo

ctt

etti

08 C. Chollet et al. / Colloids and Surfa

Another promising approach to improve implant biointegrations the realisation of micro-patterns on the biomaterial surfaces.ndeed, the presence of microstructures on a surface allows theontrol of cell–substrate interactions by creating, in some cases,atterns of cells that are highly oriented and differentiated. Forhis reason, microfabrication techniques have been widely usedo produce surface topographies to investigate the processes ofell adhesion, spreading, migration, proliferation, and gene expres-ion, as they allow precise production of specified topographies onaterial surfaces [15,16]. Indeed, a cell carrier with a controlledicrostructure would better maintain cell morphology, differenti-

tion, and functionality over long periods of time. In recent years,D micro-patterning of adhesive and non-adhesive regions waschieved [17] and used to promote selective osteoblast adhesion18], to investigate its subsequent effect on matrix mineraliza-ion and to assess the effect of the implant’s surface chemistry onsteoblast–material interactions [19]. These studies showed thatsteoblasts could be confined within patterns with an appropri-te dimension and surface chemistry. It was demonstrated thatsteogenic cells from rat calvaria showed higher ALP activitynd formed more bone nodules on Ti or hydroxyapatite-coated,icrogrooved silicon wafers than on unpatterned surfaces, and the

umber of bone nodules increased with groove depth (10–30 �m)15]. Rat bone marrow-derived osteoblasts were shown to haveigher ALP activity and mineralization on microgrooved PLA sur-

aces than on the unpatterned ones [20]. It is thus shown thaturface topography has a role in osteoblast alignment, prolifera-ion, and differentiation, but the parameters contributing to thesend the optimum values are not yet known.

These studies are very important because cell micro-patterningas broad applications in tissue engineering. Cellular activities areighly dependent on their microenvironments, such as their scaf-

olds and surrounding cells. Therefore, it is important to design andngineer the cellular microenvironment so as to make cells in vitroith functional similarities to living tissues and organs [21].

Today, there exist several methodologies that permit to gener-te biological micro-patterns on materials but most of the studiessed soft lithography techniques, such as micro-contact printingMCP) [22], microfluidic patterning [23] or micromolding [24], with

CP being by far the most widely used technique to create bio-ogical micro-patterns. Thanks to this technique, there is a wideange of approaches to directly pattern materials with cell adhesionromoting molecules. The list of applicable adhesion-promotingolecules is long. In most cases, different extra-cellular matrix

ECM) proteins or synthetic peptide constructs with ECM bindingites have been printed. For example, Bernard et al. stamped dif-erent model proteins such as immunoglobulins, BSA, and NgCAM25]. In the same way, endothelial and several other cell types haveeen patterned using fibronectin designs [26]. Soft lithographicechniques have always used self-assembled monolayers (SAMs)o immobilize biomolecules on the material surfaces, and there-ore, the molecules are not covalently grafted onto biomaterial.onsequently, the stability of the molecular patterns that are notovalently bound to the underlying substrate may not be suffi-ient for longer term cell-biological studies [27]. Moreover, SAMsenerally use alkanethiolate which presents a limited stability toxidation [28].

In this article we propose the covalent grafting of RGD-ontaining peptides onto a PET surface in well-defined microstruc-ures in order to control MC3T3 cells adhesion. To this end, we usedwo different techniques in obtaining micro-patterned materials:

(i) The first relies on a photoablation technique combiningxcimer laser with a non-contact microphotolithographical projec-ion technique in order to microfabricate well-defined surfaces, (ii)he second is a photolithography process which typically consistsn the transfer of a specific pattern to a photosensitive material (a

Biointerfaces 75 (2010) 107–114

photoresist) by selective exposure to a radiation source. Photoresistmay also be used as a template for patterning a material depositedafter lithography. The resist is subsequently etched away, and thematerial deposited on the resist is “lifted off”.

The aim of this paper is twofold: (1) to validate the feasibility anddevelopment of well-defined and reproducible covalently boundmicro-patterns and (2) to evaluate the influence of this peptide’sdistribution on the cellular behavior by using the MC3T3 osteoblastcell model.

2. Materials and methods

2.1. Materials

The PET material used was a commercial film (Mylar® D)obtained from Dupont Teijin, France. It is in the form of bi-orientedfilm with a thickness of 75 �m. Inorganic reagents (NaOH, KMnO4,H2SO4, HCl) were obtained by Sigma, France.

Acetonitrile, Dimethylaminopropyl-3-ethylcarbodiimidehydrochloride (EDC), N-hydroxysuccinimide (NHS) and (2-(N-morpholino)-ethanesulfonic acid (MES buffer) were obtainedfrom Aldrich, France. RGDC peptides were obtained from Bachem,France. S1818 Positive tone photoresist and Microposit Developerwere obtained from Shipley, France. Antibodies were provided byMolecular Probes and the DAPI (4,6-Diamidina-2-phenylin) fromSigma, France.

2.2. Methods

2.2.1. RGD peptide grafting onto poly-(ethylene terephthalate)2.2.1.1. Surface preparation of PET and RGDC grafting. The PET filmwas subjected to an ethanol washing and then sonification forpurification before use. The three subsequent modification stepsare detailed elsewhere [29]. Briefly:

1. Materials were modified by means of hydrolysis and oxidationin order to create COOH functions on the PET surface.

2. The PET-COOH material was then immersed in a solution of EDC(0.2 M) + NHS (0.1 M) + MES buffer (0.1 M) in MilliQ Water.

3. The immobilization of cell-binding peptides containing the RGDsequence was realised in a solution of RGDC for 15 h at room tem-perature. After grafting, the disks were rinsed in MilliQ Waterduring 1 week.

2.2.1.2. Surface characterization. In order to validate the chemicalmodifications at each step of the treatment, the PET surfaces werecharacterized by different physicochemical tools such as X-Rayphotoelectron spectroscopy, Toluidine Blue staining and High Res-olution �-imager. These results were described in previous works[30].

Brewster angle microscopy (BAM) measurements. The RGDC layergrafted onto the polymer was evaluated using a Brewster anglemicroscope (NFT BAM2plus, Göttingen, Germany) mounted on theLangmuir trough. The microscope was equipped with a frequencydoubled Nd:Yag laser (532 nm, 50 mW), polarizer, analyser and aCCD camera. The exposure time (ET), depending on the image lumi-nosity, was adjusted to avoid saturation of the camera. The spatialresolution of the Brewster angle microscope was about 2 �m, andthe image size was 600 �m × 450 �m with 10× magnification lensused. The BAM images are coded in gray level. To determine the

thickness of the layer at the surface, we used the calibration pro-cedure of the BAM software that determines the linear functionbetween the reflectance and the gray level. This function is estab-lished by comparison between the experimental curve of the graylevel as a function of the incidence angle and the Fresnel curve

Page 3

C. Chollet et al. / Colloids and Surfaces B:

Fet

(cFaofittBcw

22R2ibbetosdtndw

2spct2

ig. 1. Schematic of the experimental setup. The image of the mask is projected andtched onto the surface of the sample (S) after the KrF laser beam travels throughhe homogenizer (H), the field lens (F), the mask (M) and the projection lens (P).

curve of the reflectance as a function of the incidence angle) thatan be fitted by a parabola around the Brewster angle minimum.rom the reflectance value, the BAM thickness model allows evalu-tion of the thickness of the layer at the surface with the knowledgef the experimental Brewster angle and the optical index of thelm. This model is based on the proportionality relation betweenhe reflectance and the square of the interfacial film thickness whenhe optical index of the film is assumed constant. Moreover, withrewster angle microscopy, information on the fluidity of the filman be obtained by observing the geometry of the domains at theater surface.

.2.2. Elaboration of surfaces with well-defined topochemistry

.2.2.1. UV excimer laser ablation. Ablation was performed ontoGD-grafted surface by using a pulsed KrF excimer laser (248 nm,0 ns per pulse) whose beam illuminates a metallic mask consist-

ng of a slot of 750-�m width which is imaged on the graft surfacey means of a precision lens (fully aberration-corrected, designedy F. Goodall, RAL, UK) of 0.2 numerical aperture, capable of a lat-ral resolution of ∼0.5 �m (Fig. 1) [31]. The image of the mask isransferred (without contact) to the sample with a size reductionf 10, at fluence adjusted in order to achieve spontaneous ablationince threshold is of the order of ∼35 mJ/cm2. The sample was irra-iated in a step-and-repeat mode by lateral displacement adjustedo the desired distance (e.g. 5–300 �m), in order to perform theext microgroove (one cycle). The entire operation (10 cycles) wasriven manually. A series of microgrooves could thus be obtainedith 10 pulses per groove [32].

.2.2.2. Photolithography. Photosensible resin was homogeneously

pin-coated onto the PET (1 �m thick) and then dried. A photoresistattern was created by first placing a mask containing a lithographi-ally defined metal grid pattern in contact with the photoresist film,he latter being exposed to UV light filtered through the mask (15 s,20 V, 60 W). The pattern was completed by immersing the sub-

Biointerfaces 75 (2010) 107–114 109

strate in a commercially available developer solution (MicropositDeveloper (50% in water)), which removed the exposed photore-sist and left a photoresist pattern surrounded by bare substrate. PETmaterials were further dried and NHS molecules were covalentlybound to the exposed regions of the surface. Then, RGDC peptideswere covalently grafted onto the polymer surface. The photore-sist pattern was then removed by a 1 min submersion in acetonerevealing the underlying substrate [29].

2.2.2.3. Micro-patterns characterization. All micro-patterns werecharacterized by scanning electron microscopy, optical profilom-etry and high resolution micro-imager. Results have already beenpresented elsewhere [29].

2.2.3. Evaluation of a biological effect on homogeneous orpatterned surfaces2.2.3.1. Cell culture. MC3T3-E1 (osteoblast-like) cells are a non-transformed cell line established from newborn mouse calvariaand exhibit an osteoblastic phenotype. Cells were grown in Alpha-MEM medium supplemented with 10% FBS (Fetal Bovine Serum),1% penicillin/streptomycin and 1% l-glutamine. Cells were subcul-tured once a week using trypsin/EDTA and maintained at 37 ◦C in ahumidified atmosphere of 5% CO2 in air.

2.2.3.2. Cell attachment. Functionalized polymer disks were steril-ized in 70% ethanol for 10 min and then rinsed three times in PBS.They were placed in 24-well tissue culture plates (Corning) that hadbeen previously coated with 2% agarose. MC3T3 cells were seededat a density of 100,000 cells/cm2 on each material and allowed toattach for 15 h at 37 ◦C in serum-free �-MEM medium. The posi-tive and negative controls of cell adhesion were the tissue culturepolystyrene (TCPS) and agarose (2%) coating, respectively.

2.2.3.3. Morphological study. Optical microscope. A Zeiss opticalmicroscope (Axiovert) equipped with an Olympus camera was usedto observe the cells and patterns obtained by photolithography.

Fluorescent staining: F-actin filaments and DNA. Fluorescentstaining was performed 15 h after cell seeding. Cell-loaded mate-rials were rinsed twice with serum-free medium, fixed with 4%(w/v) paraformaldehyde in phosphate-buffered saline (PBS) dur-ing 15 min at 4 ◦C, and permeabilized with permeabilizing bufferpH = 7.2 (10.3% (w/v) sucrose + 0.292% (w/v) NaCl + 0.06% (w/v)MgCl2 + 0.476% (w/v) Hepes buffer + 0.5% (v/v) Triton-X-100) for15 min at 4 ◦C.

Cell cytoskeletal filamentous actin (F-actin) was visualized bytreating the cells with 5 U/mL Alexa Fluor® 488 phalloidin for 1 hat 37 ◦C. Cell nuclei were counterstained with 20 �g/ml DAPI (4,6-diamidina-2-phenylin, Sigma) during 10 min at room temperature.Samples were finally mounted in Vectashield®.

Scanning electron microscopy. Cell morphology was assessed 15 hafter cell seeding, by SEM (15 kV using a Hitachi S-2500). Cell-loaded materials were rinsed twice with PBS, fixed with 2.5% (v/v)glutaraldehyde in 0.1 M cacodylic acid buffer (pH = 7.4), and dehy-drated in serial-diluted ethanol solutions of 30, 50, 70, 80, 90, 95 and100% (v/v). Finally, they were coated by HMDS and dried under vac-uum for 2 h. Prior to observing the samples, materials were coatedwith a sputter coater equipped with a gold target.

3. Results

3.1. Surface characterization

3.1.1. BAM measurementsAt the Brewster Angle (60.2◦), BAM micrographs show a homo-

geneous layer where reflectance increases with compression from

Page 4

110 C. Chollet et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 107–114

Fig. 2. Pictures obtained by optical microscopy of MC3T3 cells 15 h after seeding in serum free medium on: (a) native PET and (b) PET-RGDC.

15 h a

Rt

3

3

gtl

3cRot

3o

Fw

Fig. 3. Pictures obtained by scanning electron microscopy of MC3T3 cells

f = 5.5 × 10−6. Estimations of RGDC layer thickness taking a refrac-ive index of 1.7 give ≈15 Å.

.2. Biological evaluation

.2.1. Homogeneous surfacesCell attachment tests were carried out on homogeneous surfaces

rafted with RGD peptides. The potential of these RGDC peptideso promote cell adhesion to the PET was investigated by morpho-ogical study using mouse osteoblastic cells (MC3T3).

.2.1.1. Optical microscopy. Optical microscope pictures of MC3T3ells 15 h after seeding onto native PET (Fig. 2(a)) and onto PET-GDC (Fig. 2(b)) show a clear difference of morphology. Cells seeded

nto native PET stay round on the surface whereas cells seeded ontohe PET-RGDC material were well attached and spread out.

.2.1.2. Scanning electron microscopy. In order to supplement theptical microscopy study, we used SEM to observe the MC3T3

ig. 4. Pictures obtained by optical microscopy of MC3T3 15 h after seeding in serum freeidth (�m): (a) 100–100 and (b) 100–50.

fter seeding in serum free medium on: (a) native PET and, (b) PET-RGDC.

cells seeded onto the native polymer (Fig. 3(a)) and onto theRGDC-modified material (Fig. 3(b)). These results were perfectlyin agreement to those obtained previously by optical microscopy.In the case of the native PET (Fig. 3(a)), we observed a retractationof the cell, which seems to indicate that it does not like the hostsurface. In opposition to this, cells seeded onto the RGD-modifiedpolymer were well spread out (Fig. 3(b)). Moreover, cells seededonto the RGD-modified material appeared to show secretions onthe surface of their membrane suggesting some kind of metabolicactivity (Fig. 3(b)).

Morphological studies of cells seeded onto homogeneous sur-faces, either modified or not by RGD, proved the efficacy of theseRGDC peptides in promoting cell adhesion.

3.2.2. Patterned surfacesThe second approach of this study consisted of the seeding of

MC3T3 cells onto a patterned surface in order to validate the effi-ciency of these two methods in creating RGD micro-patterns andin studying their effects on cell adhesion.

medium onto micro-patterns created by photolithography with a “non-RGD–RGD”

Page 5

C. Chollet et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 107–114 111

F seruc (b) 10i

3woocti

smar1ab

3o(w1t

F2

ig. 5. Pictures obtained by fluorescence microscopy of MC3T3 15 h after seeding inreated by photolithography with a “non-RGD–RGD” width (�m): (a) 100–100 ands referred to the web version of the article.)

.2.2.1. Using photolithography. Optical microscopy. Micro-patternsith various widths were elaborated (“non-RGD-RGD” width (�m)

f (a) 100–100, (b) 100–50). Fig. 4 shows, after a 15 h seeding periodnto the materials’ surface, an obvious alignment of the MC3T3ells onto the RGD lines, whatever the pattern size. Nevertheless,he lower the width of the RGD line, the more the cell lengthensnto the microgrooves.

Fluorescent staining: F-actin filaments and DNA. Fluorescenttaining was used to confirm results obtained by opticalicroscopy. We chose to stain F-actin filaments and DNA to have

n idea of cell organisation onto the micro-patterned surface. Fig. 5epresents pictures obtained by fluorescence microscopy of cells5 h after seeding onto RGD-patterned PET. We can observe anlignment of cells in RGD grooves and the ability of cells to bindetween them.

.2.2.2. Using UV excimer laser ablation. Optical microscope. The

ptical micrographs of four types of laser microfabricated surfacesa–d) presenting micro-patterns with a “RGD width”–“ablationidth”–“ablation depth” (�m) of (a) 25–75–1.2, (b) 85–75–1.2, (c)

25–75–1.2 and (d) 325–75–1.2, after a 15 h-seeding period withhe MC3T3 cells, are shown in Fig. 6(a–d). Whatever the size of

ig. 6. Pictures obtained by optical microscopy of MC3T3 15 h after seeding onto micro5–75–1.2, (b) 85–75–1.2, (c) 125–75–1.2 and (d) 325–75–1.2.

m free medium (F-actin filaments in green and nuclei in blue) onto micro-patterns0–50. (For interpretation of the references to color in this figure legend, the reader

the pattern carried out, the cells seem to align themselves into theRGD-functionalized area.

Scanning electron microscopy. The scanning electron micro-graphs of micro-patterns with a RGD width of 5 �m and 85 �m,after a 15 h-seeding period with the MC3T3 cells, are respectivelyshown in Fig. 7(a) and (b–c). In all pictures, we can observe thatcells are aligned on the RGD patterns and not on ablated areas.That is perfectly expressed in Fig. 7(a) where we can notice the cellperfectly aligned on the 5 �m RGD area. Fig. 7(a) and (c) showsthe orientation of the cells on the inter-ablated spacing, with thelatter appearing to show the cells connecting by the extension ofpseudopods towards each other.

4. Discussion

4.1. Photolithography

Photolithography is a technique currently used in electronicmanufacturing for its low cost and its wide range of designs.However, this technique has quickly been adapted for biologicalapplications. For example, Veishe et al. used photolithography toimmobilize proteins on gold surfaces in order to enhance cell adhe-

-patterned surfaces with a RGD width–ablation width–ablation depth (�m); (a)

Page 6

112 C. Chollet et al. / Colloids and Surfaces B: Biointerfaces 75 (2010) 107–114

F ing on

sstulp

pdb

jpsImsoroticstTtav

4

pttrtso(

aFpoof

ig. 7. Pictures obtained by scanning electron microscopy of MC3T3 15 h after seed

ion [33]. In this study, they proved the ability of the engineeredurfaces to guide cell adhesion. However, the use of solvents duringhe lift-off process can offer a disadvantage when proteins are beingsed, due to potential protein denaturation and consequently the

oss of their biological activity. A solution can then be to use shorteptides which are more stable than proteins [34].

In this paper, we chose to covalently graft RGDC peptides ontohotolithographic micro-patterns presenting two different specificesigns, with the ability of creating numerous other designs havingeen described previously [29].

As described in previous paper [29], patterns are controlledust after resin removal by optical microscope and then by opticalrofilometer. Briefly, pictures obtained with these two techniqueshow that patterns are well defined and are close to mask designs.ndeed, domain sizes and designs are similar to those present on the

ask which prove that UV irradiation is perfectly controlled. Theecond step consists of evaluating the cell attachment onto patternsbtained by photolithography. Figs. 4 and 5 show pictures, obtainedespectively by optical microscope and by fluorescence microscope,f MC3T3 15 h after seeding on RGDC-patterned polymer. Whateverhe size of the pattern, cells appeared to be connecting by extend-ng pseudopods towards each other. We can see an alignment ofells onto the RGDC lines, nevertheless cells seem more likely topread out when they adhere onto larger RGDC domains. However,he lower the RGDC groove width, the greater the cell orientation.his result is in agreement with those previously demonstrated inhe literature; Clark et al. proved that when the grooves or ridgesre appreciably wider than the cells, effects on orientation are notery marked [35].

.2. UV excimer laser ablation

UV excimer laser ablation is another way to create micro-atterns onto RGDC-grafted PET. This method does not requirehe use of organic solvents that preserve the integrity of the pep-ides. However, we have to ensure that all RGDC peptide layers areemoved during ablation. Thus, ablation depth was determined inhe function of pulse numbers by optical profilometer. As we canee in previous studies [29], with 10 pulses, we ablated a thicknessf 1.5 �m, a thickness which is sufficient to remove all peptidesRGD layer = 1.5 nm obtained by BAM measurements).

In this study, MC3T3 cells were seeded onto ablated materialsnd observed by optical microscopy and SEM after 15 h of culture.

ig. 7 shows pictures representing cells seeded on three differentatterns. In all cases, we can see that cells are exclusively attachednto the RGDC areas. Moreover, cell spreading seems dependentn the width of the RGD layer. Indeed, the best alignment occursor an RGDC width ranging from 25 �m to 125 �m.

to micro-patterned surfaces with a RGD spacing (�m); (a) 5, (b) 85, and (c) 85.

However, laser treatment of PET has physical and chemical con-sequences which have been studied for a few years [36,37]. Forexample, laser ablation is able to change the surface propertiesof the material and consequently modify platelet adhesion andcell behavior [38,39]. It subsequently appears predictable that thephysicochemical modifications of the surface will influence theinteractions between the cells and the substrate. Numerous studiesproved the effect of roughness [40,41] and hydrophilicity [42,43]of material surface on cell behavior.

It is already known that laser ablation creates various kinds ofoxidized groups, caused by the decomposition of peroxides by laserpulses, on the PET surface [38], and these groups are known to mod-ify cell adhesion [44,45]. Moreover, laser ablation modifies materialtopography creating microgrooves on the surface that can affect cellresponse [46,47]. Duncan et al. used laser excimer beam technologyto create PET surfaces with well-defined 3D microdomains in orderto study Human Osteoblast Progenitor (HOP) cell growth [48]. Theyproved that HOP adhered preferentially on surfaces with “smooth”microtopographical transitions (i.e. minimal microgroove widthand depth). These results show that the presence of grooves inhibitsthe cellular adhesion but it is very difficult to dissociate this effectfrom the influence of the chemical modifications on the cellularadhesion.

In our study, a certain question arised: is cell alignment dueto RGDC grafting or to chemical and topographic modificationsinduced by laser? Britland et al. [49] studied this question usinga chemical cue (laminin) oriented at right angles to a topographicone. When the grooves were 500 nm deep or less, the cells reactedchiefly to the chemical cue. On deeper grooves (5 �m depth) thetopographic effect oriented about 80% of cells and the chemicalone 7%. This result confirmed those obtained by Clark et al. [35].

In our study, groove depths were less than 1.5 �m and, thus,would not be sufficient to orient cell adhesion, agreeing with theresults obtained by Britland et al. and by Clark et al., even if the cellsthey used were different. Studying BHK (Baby Hamster Kidney),MDCK (Madin-Darby Canine Kidney) and chick embryo cerebralneurones, they proved that cell behavior on patterned surfaces canbe different in function of the cell line used.

Moreover, these results are similar with those obtained usingphotolithography. These results seem sufficient to conclude thatcell alignment on modified PET is due to the presence of RGD pep-tides which enhance cell adhesion.

In all cases, biological results are the same: cells exhibit an align-ment on RGD areas and appear to be connecting, whatever the

space between RGD domains. However, the size of pattern (widthof RGD layer) seems to be a very important parameter. Indeed,Figs. 4 and 5 prove, in the case of patterns obtained by photolithog-raphy, that cell spreading is better with a 100 �m RGD width. Thisresult is confirmed in Figs. 6 and 7. Indeed, in the case of pat-

Page 7

ces B:

ttusisd

5

RdipabsOctfbscolcpoeteg

A

MU3

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

C. Chollet et al. / Colloids and Surfa

erns created by photoablation the best alignment occurs for RGDhickness of 85 �m. However, these results are valid only with these of MC3T3 cells, because cell response onto micro-patternedurfaces depends clearly on cell type size. Indeed, Meyle et al. stud-ed cell growth of a variety of cell types on micro-textured silicaurfaces and showed the effect of cell orientation to be cell typeependent [50].

. Conclusion

A three-step reaction procedure was developed to attachGD-containing peptides onto a PET surface with a controlledistribution in order to clarify the capacity of RGD peptides to

nteract with integrin receptors to promote cell adhesion. Micro-atterns were created onto a polymer surface using two differentpproaches: photolithography and UV excimer ablation laser. Theiological evaluation of micro-patterned surfaces was done bytudying adhesion of MC3T3 in serum-free conditions during 15 h.ptical microscopy, scanning electron microscopy and fluores-ence microscopy were used to observe cell morphology. Whateverhe technique used to create micro-patterns, a cell alignment occursor RGD thickness around 100 �m. This work will be completedy mineralization studies and by elaboration of bi-functionalizedurfaces able to allow adhesion of endothelial cells and osteoblastells. Bone growth and remodeling depend on complex interactionsccurring between osteoblasts and other cells present, particu-arly vascular endothelial cells that may be pivotal members of aomplex interactive communication network in the bone. Futurerospects would deal with the realisation of a surface grafted,n one hand, with an osteoblast specific growth factor known tonhance osteoblast mineralization (for example BMP2) and, onhe other hand, with REDV peptides able to specifically increasendothelial cell adhesion in order to improve material osteointe-ration.

cknowledgement

The authors would like to thank B. Desbat for Brewster Angleicroscopy measurements (Institut des Sciences Moléculaires,MR 5255, Université Bordeaux 1, 351 cours de la Libération -3405 Talence cedex France).

eferences

[1] R. Mikulikova, et al., Cell microarrays on photochemically modified polyte-trafluoroethylene, Biomaterials 26 (27) (2005) 5572–5580.

[2] N. Blanchemain, et al., Vascular PET prostheses surface modification withcyclodextrin coating: development of a new drug delivery system, EuropeanJournal of Vascular and Endovascular Surgery 29 (6) (2005) 628–632.

[3] H. Gappa-Fahlenkamp, R.S. Lewis, Improved hemocompatibility ofpoly(ethylene terephthalate) modified with various thiol-containing groups,Biomaterials 26 (17) (2005) 3479–3485.

[4] A.V. Vasil’ev, et al., Repair of corneal defects using tissue engineering, Izv. Akad.Nauk. Ser. Biol. 1 (2005) 5–8.

[5] J.G. Meinhart, et al., Clinical autologous in vitro endothelialization of 153infrainguinal ePTFE grafts, Ann. Thorac. Surg. 71 (5 Suppl.) (2001) S327–S331.

[6] U. Kneser, et al., Tissue engineering of bone: the reconstructive surgeon’s pointof view, J. Cell. Mol. Med. 10 (1) (2006) 7–19.

[7] Z. Zhang, et al., Neurite outgrowth on well-characterized surfaces: preparationand characterization of chemically and spatially controlled fibronectin and RGDsubstrates with good bioactivity, Biomaterials 26 (1) (2005) 47–61.

[8] R.O. Hynes, Integrins: bidirectional, allosteric signaling machines, Cell 110 (6)(2002) 673–687.

[9] M.D. Pierschbacher, E. Ruoslahti, Cell attachment activity of fibronectin can beduplicated by small synthetic fragments of the molecule, Nature 309 (5963)(1984) 30–33.

10] M.C. Porté-Durrieu, et al., Development of RGD peptides grafted onto silica sur-faces: XPS characterization and human endothelial cell interactions, J. Biomed.Mater. Res. 46 (3) (1999) 368–375.

11] M.C. Porte-Durrieu, et al., Cyclo-(DfKRG) peptide grafting onto Ti–6Al–4V:physical characterization and interest towards human osteoprogenitor cellsadhesion, Biomaterials 25 (19) (2004) 4837–4846.

[

[

Biointerfaces 75 (2010) 107–114 113

12] K.E. Michael, et al., Adsorption-induced conformational changes in fibronectindue to interactions with well-defined surface chemistries, Langmuir 19 (19)(2003) 8033–8040.

13] K.E. Healy, et al., Kinetics of bone cell organization and mineralizationon materials with patterned surface chemistry, Biomaterials 17 (2) (1996)195–208.

14] G. Maheshwari, G. Brown, D.A. Lauffenburger, A. Wells, L.G. Griffith, Cell adhe-sion and motility depend on nanoscale RGD clustering, J. Cell Sci. 113 (2000)1677–1686.

15] D. Perizzolo, W.R. Lacefield, D.M. Brunette, Interaction between topographyand coating in the formation of bone nodules in culture for hydroxyapatite- andtitanium-coated micromachined surfaces, J. Biomed. Mater. Res. 56 (4) (2001)494–503.

16] D.S. Vara, et al., Cardiovascular tissue engineering: state of the art, Pathol. Biol.53 (10) (2005) 599–612.

17] D. Lehnert, et al., Cell behaviour on micropatterned substrata: limits of extra-cellular matrix geometry for spreading and adhesion, J. Cell Sci. (2004) 41–52.

18] M.E. Hasenbein, T.T. Andersen, R. Bizios, Micropatterned surfaces modified withselect peptides promote exclusive interactions with osteoblasts, Biomaterials23 (19) (2002) 3937–3942.

19] H. Kenar, G.T. Köse, V. Hasirci, Tissue engineering of bone on micropatternedbiodegradable polyester films, Biomaterials 27 (6) (2006) 885–895.

20] K. Matsuzaka, et al., The effect of poly-lactic acid with parallel surfacemicro groove on osteoblast-like cells in vitro, Biomaterials 20 (14) (1999)1293–1301.

21] J. Nakanishi, et al., Recent advances in cell micropatterning techniques for bio-analytical and biomedical sciences, Anal. Sci. 24 (1) (2008) 67–72.

22] T.T. Huang, T. Geng, D. Akin, W.-J. Chang, J. Sturgis, R. Bashir, A.K. Bhu-nia, J.P. Robinson, M.R. Ladisch, Micro-assembly of functionalized particulatemonolayer on C18 derivatized SiO surfaces, Biotechnol. Bioeng. 83 (4) (2003)416–427.

23] J.S. Miller, et al., Laser-scanning lithography (LSL) for the soft lithographic pat-terning of cell-adhesive self-assembled monolayers, Biotechnol. Bioeng. 93 (6)(2006) 1060–1068.

24] A. Khademhosseini, et al., Micromolding of photocrosslinkable hyaluronic acidfor cell encapsulation and entrapment, J. Biomed. Mater. Res. A 79 (3) (2006)522–532.

25] A. Bernard, et al., Printing patterns of proteins, Langmuir 14 (9) (1998)2225–2229.

26] C.D. James, et al., Patterned protein layers on solid substrates by thin stampmicrocontact printing, Langmuir 14 (4) (1998) 741–744.

27] D. Choquet, D.P. Felsenfeld, M.P. Sheetz, Extracellular matrix rigidity causesstrengthening of integrin-cytoskeleton linkages, Cell 88 (1) (1997) 39–48.

28] D. Falconnet, et al., Surface engineering approaches to micropattern surfacesfor cell-based assays, Biomaterials 27 (16) (2006) 3044–3063.

29] C. Chollet, et al., RGD peptides micro-patterning on poly(ethylene terephtha-late) surfaces, ITBM-RBM 28 (1) (2007) 2–12.

30] C. Chollet, et al., RGD peptides grafting onto poly(ethylene terephthalate) withwell controlled densities, Biomol. Eng. 24 (5) (2007) 477–482.

31] F. Weisbuch, et al., Millimeter long PMMA nanofibers—a new form of materialremoval in laser ablation, Thin Solid Films 453–454 (2004) 394–398.

32] C. Chollet, et al., RGD peptide grafting onto micro-patterned PET: peptide dis-tribution impact on cell attachment, J. Laser micro/nano Eng. 1 (3) (2006)226–230.

33] M. Veiseh, et al., Guided cell patterning on gold-silicon dioxide substrates bysurface molecular engineering, Biomaterials 25 (16) (2004) 3315–3324.

34] Y. Ito, et al., Culture of human umbilical vein endothelial cells on immobi-lized vascular endothelial growth factor, J. Biomed. Mater. Res. A 74 (4) (2005)659–665.

35] P. Clark, et al., Topographical control of cell behaviour: II. Multiple groovedsubstrata, Development 108 (4) (1990) 635–644.

36] A.C. Duncan, S. Lazare, C. Baquey, Use of biological nanoprobes for the charac-terization of micropatterned surfaces obtained by ultraviolet laser lithography,Appl. Surf. Sci. 221 (1–4) (2004) 93–98.

37] G. Mayer, et al., Physico-chemical and biological evaluation of excimer laserirradiated polyethylene terephthalate (pet) surfaces, Biomaterials 27 (4) (2006)553–566.

38] M. Dadsetan, et al., Cell behavior on laser surface-modified polyethyleneterephthalate in vitro, J. Biomed. Mater. Res. 57 (2) (2001) 183–189.

39] M. Dadsetan, et al., In vitro studies of platelet adhesion on laser-treatedpolyethylene terephthalate surface, J. Biomed. Mater. Res. 54 (4) (2001)540–546.

40] D.W. Hamilton, D.M. Brunette, The effect of substratum topography onosteoblast adhesion mediated signal transduction and phosphorylation, Bio-materials 28 (10) (2007) 1806–1819.

41] I. Degasne, et al., Effects of roughness, fibronectin and vitronectin on attach-ment, spreading, and proliferation of human osteoblast-like cells (Saos-2) ontitanium surfaces, Calcified Tissue Int. 64 (6) (1999) 499–507.

42] K. Webb, V. Haldy, P.A. Tresco, Relative importance of surface wettability andcharged functional groups on NIH 3T3 fibroblast attachment, spreading, and

cytoskeletal organization, J. Biomed. Mater. Res. 41 (3) (1998) 422–430.

43] J.L. Dewez, et al., Influence of substrate hydrophobicity on the adsorption of col-lagen in the presence of pluronic F68, albumin, or calf serum, J. Colloid InterfaceSci. 191 (1) (1997) 1–10.

44] N. Faucheux, et al., The dependence of fibrillar adhesions in human fibroblastson substratum chemistry, Biomaterials 27 (2) (2006) 234–245.

Page 8

1 ces B:

[

[

[

[

14 C. Chollet et al. / Colloids and Surfa

45] B.G. Keselowsky, D.M. Collard, A.J. Garcia, Surface chemistry modulates focal

adhesion composition and signaling through changes in integrin binding, Bio-materials 25 (28) (2004) 5947–5954.

46] A. Curtis, C. Wilkinson, Topographical control of cells, Biomaterials 18 (24)(1997) 1573–1583.

47] K. Anselme, M. Bigerelle, Topography effects of pure titanium substrates onhuman osteoblast long-term adhesion, Acta Biomater. 1 (2) (2005) 211–222.

[

[

Biointerfaces 75 (2010) 107–114

48] A.C. Duncan, et al., Effect of laser modified surface microtopochemistry

on endothelial cell growth, Colloids Surf. B: Biointerfaces 54 (2) (2007)150–159.

49] S. Britland, et al., Morphogenetic guidance cues can interact synergistically andhierarchically in steering nerve cell growth, Exp. Biol. Online 1 (2) (1997) 1–15.

50] J. Meyle, K. Gültig, W. Nisch, Variation in contact guidance by human cells on amicrostructured surface, J. Biomed. Mater. Res. 29 (1) (1995) 81–88.