Surface Investigation on Biomimetic Materials to Control Cell

Mar 29, 2010 - In biological tissues, cells are immersed in the extracellular matrix (ECM) that is a coacervate of glycosaminoglycans and proteins wit...
54 downloads 11 Views 1023KB Size
pubs.acs.org/Langmuir © 2010 American Chemical Society

Surface Investigation on Biomimetic Materials to Control Cell Adhesion: The Case of RGD Conjugation on PCL Filippo Causa,* Edmondo Battista, Raffaella Della Moglie, Daniela Guarnieri, Maria Iannone, and Paolo A. Netti Interdisciplinary Research Centre on Biomaterials (CRIB) University Federico II, Piazzale Tecchio 80, 80125, Naples, Italy, and Italian Institute of Technology (IIT) Via Morego, 30 Genoa, Italy Received January 15, 2010. Revised Manuscript Received March 15, 2010 The cell recognition of bioactive ligands immobilized on polymeric surfaces is strongly dependent on ligand presentation at the cell/material interface. While small peptide sequences such as Arg-Gly-Asp (RGD) are being widely used to obtain biomimetic interfaces, surface characteristics after immobilization as well as presentation of such ligands to cell receptors deserve more detailed investigation. Here, we immobilized an RGD-based sequence on poly(ε-caprolactone) (PCL), a largely widespread polymeric material used in biomedical applications, after polymer aminolysis. The surface characteristics along with the efficacy of the functionalization was monitored by surface analysis (FTIR-ATR, contact angle measurements, surface free energy determination) and spectrophotometric assays specially adapted for the analytical quantification of functional groups and/or peptides at the interface. Particular attention was paid to the evaluation of a number, morphology, and penetration depth of immobilized functional groups and/or peptides engrafted on polymeric substrates. In particular, a typical morphology in peptide distribution was evidenced on the surface raised from polymer crystallites, while a significant penetration depth of the engrafted molecules was revealed. NIH3T3 fibroblast adhesion studies verified the correct presentation of the ligand with enhanced cell attachment after peptide conjugation. Such work proposes a morphological and analytical approach in surface characterization to study the surface treatment and the distribution of ligands immobilized on polymeric substrates.

Introduction In biological tissues, cells are immersed in the extracellular matrix (ECM) that is a coacervate of glycosaminoglycans and proteins with various mechanical and signaling functions. In particular, fibroblast and osteoblast cells are known to express various integrins, each component having a large extracellular domain responsible for ligand binding, a transmembrane domain, and a short cytoplasmic domain responsible for interacting with the actin cytoskeleton.1 Integrin heterodimers bind to specific amino acid sequences, such as the arginine-glycine-aspartic acid (Arg-Gly-Asp or RGD) recognition motif that is largely present in many ECM proteins, including fibronectin, vitronectin, bone sialoprotein, and osteopontin.2 Small synthetic peptides (a few hundred daltons) that contain amino acid sequence RGD can thus mediate cell attachment as well as the large parental molecule (a hundred thousand dalton). On the basis of this, biomimetic approaches have been developed to immobilize short peptides, such as RGD, onto synthetic or natural surfaces, to produce biofunctional materials able to promote and enhance cell attachment.1,3 In particular, it has been found that a minimum RGD density of 1.0  10-15 mol/cm2, corresponding to a spacing of about 140 nm between peptide ligands, is sufficient to promote cell spreading, while a density of 1.0  10-14 mol/cm2 is needed to promote the formation of focal contacts.4 However, such *To whom correspondence should be addressed. Interdisciplinary Research Centre on Biomaterials (CRIB), University Federico II, Piazzale Tecchio 80, 80125, Naples, Italy. Telephone number: þ39-081-7682100, fax number: þ39-081-7682404; e-mail address: [email protected].

(1) Hersel, U.; Dahmen, C.; Kessler, H. Biomaterials 2003, 24, 4385–4415. (2) Garcia, A. J.; Reyes, C. D. J. Dent. Res. 2005, 84, 407–413. (3) El-Amin, S. F.; Kofron, M. D.; Attawia, M. A.; Lu, H. H.; Tuan, R. S.; Laurencin, C. T. Clin. Orthop. Rel. Res. 2004, 427, 220–225. (4) Massia, S. P.; Hubbell, J. A. J.Cell. Biol. 1991, 114, 1089–1100.

Langmuir 2010, 26(12), 9875–9884

parameters strongly depend on peptide presentation and, in turn, from chemical and physical characteristics of the substrate. Moreover, spatial distribution and the aggregation of RGD peptides at the micro- and nanoscale significantly affect cell responses. For example, nanoscale clustering of RGD peptides can induce integrins to cluster, thus triggering complete cell signaling.5,6 Poly(ε-caprolactone) (PCL), a biodegradable aliphatic polyester,7,8 has been suggested for a wide field of applications such as drug delivery systems,9,10 tissue-engineered skin (plain film), and scaffolds for supporting fibroblast and osteoblast growth.11,12 However, as any other synthetic polymer it does not present molecular motifs for cell biological recognition, and therefore it, lacks a friendly interface with living cells.13 A way to obtain PCL biomimetic surfaces in promoting cell adhesion was the modification of polymeric backbone to introduce functional groups for the following RGD conjugation.14,15 Marletta et al. demonstrated that, when adsorbed onto PCL surface, RGD seems to have only (5) Maheshwari, G.; Brown, G.; Lauffenburger, D. A.; Wells, A.; Griffith, L. G. J. Cell. Sci. 2000, 113, 1677–1686. (6) Yang, H.; Kao, W. J. Int. J. Nanomed. 2007, 2, 89–99. (7) Eldsater, C.; Erlandsson, B.; Renstad, R. A.; Albertsson, C.; Karlsson, S. Polymer 2000, 41, 1297–1304. (8) Choi, E. J.; Kim, C. H.; Park, J. K. Macromolecules 1999, 32, 7402–7408. (9) Zhong, Z. K.; Sun, X. Z. S. Polymer 2001, 42, 6961–6969. (10) Allen, C.; Han, J.; Yu, Y.; Maysinger, D.; Eisenberg, A. J. Controlled Release 2000, 63, 275–286. (11) Ng, K. W.; Hutmacher, D. W.; Schantz, J. T.; Ng, C. S.; Too, H. P.; Lim, T. C.; Phan, T. T.; Teoh, S. H. Tissue Eng. 2001, 7, 441–455. (12) Hutmacher, D. W.; Schantz, T.; Zein, I.; Ng, K. W.; Teoh, S. H.; Tan, K. C. J. Biomed. Mater. Res. 2001, 55, 203–216. (13) Croll, T. I.; O’Connor, A. J.; Stevens, G. W.; Cooper-White, J. J. Biomacromolecules 2004, 5, 463–473. (14) Healy, K. E.; Tsai, D.; Kim, J. E. Mater. Res. Soc. Symp. Proc. 1992, 252, 109–114. (15) McConachie, A.; Newman, D.; Tucci, M.; Puckett, A.; Tsao, A.; Hughes, J.; Benghuzzi, H. Biomed. Sci. Instrum. 1999, 35, 45–50.

Published on Web 03/29/2010

DOI: 10.1021/la100207q

9875

Article

a minor effect on the cell response.16 However, both ion irradiation and RGD adsorption on PCL surfaces modulated the expression of integrins involved in human osteoblast (hOB) growth and function.17 However, the majority of these modification methods have several drawbacks including low level of functional groups, lack of control of peptide presentation on PCL surface, and the possibility of unknown and uncontrolled degradation products. Further, such methods leave the bioactive groups merely adsorbed onto the surface (i.e., they are not covalently attached), meaning that there is a danger of being exchanged or removed upon introduction into existing in vitro culture or in vivo implantation. Conversely, chemical methods were demonstrated to be successful at bioactivating polymer surface. The introduction of functional groups on biodegradable polyester surfaces and in particular on PCL has also previously been achieved by hydrolysis,13,18 aminolysis,13 plasma treatment,19,20 or copolymerization.21 Amine groups were grafted onto film surfaces through treatment with a diamine, before attaching peptide sequences such as RGD by using either carbodiimide, glutaraldehyde, or epoxy-amine chemistry.22 Subsequent cell adhesion studies have demonstrated an increase in adhesion and spreading of cells on these modified surfaces,23 in particular, on PCL containing laminin-derived peptides sequences IKVAV, RGD, or YIGSR covalently attached to the surface of the polymer. Peptides were attached to the surface of the polymer using a two-step procedure that employs a treatment with 1,6-hexanediamine followed by the use of 1-ethyl-3-(dimethylaminopropyl) carbodiimide.23 Recent work has also modified PCL with poly(ethylene oxide) grafts before coupling with RGDcontaining peptides, again resulting in enhanced cellular responses.24 PCL was also functionalized with RGD after aminolysis for three-dimensional PCL scaffold.25,26 However, in all above-mentioned works, there is a lack of control of effective presentation of immobilized peptide toward cell, and surface characteristics after peptide conjugation are not adequately investigated. Thus, a systematic study on peptide ligand organization and spatial distribution on polymer surfaces could represent a step forward in engineering bioactive interfaces and, in particular, to evaluate the effective peptide distribution and presentation able to trigger specific cell function such as adhesion or differentiation. Here, we performed the grafting of GRGDY peptide onto solid PCL with a wet chemistry consisting of a two-step procedure: polymer aminolysis to graft functional groups (primary amines) on the film surface and a following conjugation of the RGD motif. Even though the procedure is similar to others already reported in the literature,22,23 each step was controlled through functional (16) Marletta, G.; Ciapetti, G.; Satriano, C.; Pagani, S.; Baldini, N. Biomaterials 2005, 26, 4793–4803. (17) Amato, I.; Ciapetti, G.; Pagani, S.; Marletta, G.; Satriano, C.; Baldini, N.; Granchi, D. Biomaterials 2007, 28, 3668–3678. (18) Cai, K.; Yao, K.; Cui, Y.; Yang, Z.; Li, X.; Xie, H.; Qing, T.; Gao, L. Biomaterials 2002, 23, 1603–1611. (19) Yang, J.; Bei, J.; Wang, S. Biomaterials 2002, 23, 2607–2614. (20) Chu, P. K.; Chen, J. Y.; Wang, L. P.; Huang, N. Mater. Sci. Eng., R: Rep 2002, R36, 143-206. (21) Zhu, Y.; Chian, K. S.; Chan-Park, M. B.; Mhaisalkara, P. S.; Ratner, B. D. Biomaterials 2006, 27, 68–78. (22) Gabriel, M.; Van Nieuw Amerongen, G. P.; Van Hinsbergh, V. W. M.; Van Nieuw Amerongen, A. V.; Zentner, A. J. Biomater. Sci. Polym. 2006, 17, 567–577. (23) Santiago, L. Y.; Nowak, R. W.; Rubin, J. P.; Marra, K. G. Biomaterials 2006, 27, 2962–2969. (24) Taniguchi, I.; Kuhlman, W. A.; Mayes, A. M.; Griffith, L. G. Polym. Int. 2006, 55, 1385–1397. (25) Dalton, P. D.; Woodfield, T.; Hutmacher, D. W. Biomaterials 2009, 30, 701–702. (26) Lam, C. X.; Savalani, M. M.; Hutmacher, D. W. Biomed. Mater. 2008, 3, 034108.

9876 DOI: 10.1021/la100207q

Causa et al.

group determination as well as chemical and physical parameter evaluation. The aminolysis reaction was carried out for PCL surfaces and monitored. Morphological and topological characteristics of RGD presenting surfaces were deeply studied in terms of peptide surface density and distribution onto PCL surface as well as penetration depth in the polymer substrate. Finally, in order to investigate cell recognition of bioactive peptide, adhesion of NIH3T3 fibroblasts to the proposed PCL substrates was performed in serum-free media to exclude any role played by serum proteins in cell recognition.

Materials and Methods The poly(ε-caprolactone) (PCL) pellets used in this study, (Mw = 65 000 g mol-1, 181 609) are a product of Sigma-Aldrich, St. Louis, MO. The following reagents were purchased from Sigma-Aldrich, St. Louis, MO: 1,6-hexanediamine (DEA), hexylamine, glutaraldehyde solution grade I 25% (GA), glycerol, tritolyl phosphate (mixture of isomers, 90%, 268 917), sodium cyanoborohydride (NaBH3CN, Fluka, 71 435), ethanolamine, Tween 20, QuantiPro BCA assay kit (kit component: QuantiPro Buffer QA 250 mL, QuantiPro BCA QB 250 mL, 4% copper(II) sulfate pentahydrate solution 12 mL), Kaiser test kit (Fluka, 60017, the test kit contains 50 mL each of the following solutions: phenol ∼80% in ethanol, KCN in H2O/pyridin, ninhydrin 6% in ethanol), fluoresceinamine isomer I (Fluka, 07980) (FLUO), Rhodamine B isothiocyanate mixed isomers (Sigma, R1755) (RBITC), phosphate buffered saline (PBS), dichloromethane (DCM), water (CHROMASOLV Plus, for HPLC, 34877), ethanol, 2-propanol (IPA). The peptides GRGDY and GYDGR were purchased from INBIOS S.r.l., Naples, Italy. Aminolysis of Poly(ε-caprolactone) Plates. Poly(ε-caprolactone) (PCL) pellets were processed to form thin sheets by hot compression above polymer melting temperature (70 °C) for 2 h and equilibrated at room temperature. As prepared, the sheets were thoroughly washed with water and isopropanol (IPA) and flushed (or dried) with nitrogen. A procedure already described by Zhu et al.27 was followed in order to incorporate amino groups onto the surface of PCL using 1,6-hexanediamine (DEA) (Scheme 1). Briefly, aminolysis was conducted by immersing the sheets in a 10% (w/w) 1,6-hexanediamine/isopropanol (DEA/IPA) solution in a custom-made reactor thermostatted in a water bath at various temperature with adequate magnetic stirring for suitable period time. After aminolysis treatment, the samples were rinsed extensively with a 0.3% Tween 20 solution and deionized water until neutral reaction. Subsequently, the sheets were dried in a vacuum desiccator at room temperature for 24 h. In order to evaluate kinetic parameters of reaction, PCL samples were aminolyzed in 10% DEA/IPA at 25, 37, and 40 °C for 30 min. Activation energy (Ea) of aminolysis reaction was calculated according to the Arrhenius equation by fitting the plot of grafting rate against 1/T. The rate of grafting was calculated in the first 30 min from the slope of the curve at each temperature (see Supporting Information for more details). The aminolyzed PCL (PCL-NH2) plates were subsequently weighed to record mass loss and processed for physicochemical characterization. Spectroscopy after Aminolysis. ATR/FTIR measurements were performed by using a Thermo Nicolet 6700 spectrometer equipped with Smart Perfomer accessory by using a Ge crystal with incident angle of 45° and a sampling area of 2 mm. Thin films were pressed onto the Ge crystal at 12 psi (82.5 kPa) and spectra were collected in the range 4000-650 cm-1 at 8 cm-1 resolution with 128 scans. To detect at the molecular level the presence of primary amines on PCL surface, ATR/FTIR analysis was performed directly on several points of the PCL-NH2 sheets dried in a (27) Zhu, Y.; Gao, C.; Liu, X.; Shen, J. Biomacromolecules 2002, 3, 1312–1319.

Langmuir 2010, 26(12), 9875–9884

Causa et al.

Article

Scheme 1. Synthetic Scheme Showing the Two Steps Procedure Used to Immobilize Peptide Sequence on PCL Surfacea

a (1) Aminolysis of PCL substrate by diamine solutions, (2) primary amine presenting PCL substrate, (3) tether (GA) insertion and following peptide (GRGDY or GYDGR) conjugation.

vacuum desiccator without any additional treatment (n = 4 for each measurements). The penetration depth of an evanescent wave in the sample near the region more diagnostic (λ 1650 cm-1) was calculated by applying following equation:28 dp ¼

λ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2πn1 sin2 θ - ðn2 =n1 Þ2

where n2 (∼1.5) is the refractive index generally assumed for polymers, while n1 (4.0) is the refractive index of Ge crystal.28 In particular, single internal reflection element (IRE) of Ge with an incident angle (θ) of 45° allows us to reach qualitative information about samples in thin layer as high as 0.4 μm. Determination of Engrafted Amines after Aminolysis. The overall amino groups grafted onto PCL-NH2 plates were quantified by using the Kaiser test kit.29 In this assay, ninhydrin reacting with free amino groups emerging from the polymer surface produces a purple pigment (Ruhemann’s purple) detected using a spectrophotometer (Lambda 25, Perkin-Elmer) in the range 550-570 nm. The assay was conducted in homogeneous phase by dissolving the polymer in the organic component of the kit, and at the final stage of reaction, as reported by others,23,27,30 developing the pigment in an organic solvent (see Supporting Information for details). The amino density per sample area was calculated by comparing the absorbance at 551 nm with a previously obtained calibration curve (Supporting Information Figure S1).

Weight Loss, Mechanical and Thermal Properties after Aminolysis. To evaluate the effects of the aminolysis on the bulk properties of the polymer, PCL substrates of about 60 mg with an average area of 8 cm2 and rectangular shape were treated in a 10% DEA/IPA solution for a predetermined period of time ranging from 15 to 24 hs. Residual mass after aminolysis treatment was evaluated by using the gravimetric method on an electronic balance with a resolution of 0.1 mg (Gibertini, E50S). Dried plates were weighed at each time point of treatment, and data were reported as mass loss related to the projection of the treated surface area as shown in the following equation:  %RM ¼

1-

 ðMi - Mf Þ=Mi Þ  100 A

where Mi is the initial mass, Mf is the final mass, and A is the sample area. (28) Urban, M. W. Attenuated Total Reflectance Spectroscopy of Polymers Theory and Practice; American Chemical Society: Washington, DC, 1996. (29) Sarin, V. K.; Kent, S. B.; Tam, J. P.; Merrifield, R. B. Anal. Biochem. 1981, 117, 147–157. (30) Zhang, H.; Lin, C.-Y; Hollister, S. J. Biomaterials 2009, 30, 4063–4069.

Langmuir 2010, 26(12), 9875–9884

Mechanical tests were performed to assess tensile properties of dried thin sheets in accordance with the ASTM D1708-06a standard at room temperature. A set of samples with an average area of 20 cm2 and a cross section of 0.18 μm were aminolyzed in a 10% DEA/IPA solution at 37 °C and cut in dog-bone shapes as prescribed by ASTM standards. All measurements were carried out by Instron 5566 dynamometer with a 1 kN load-cell and a cross-head speed of 100 mm/min. The elastic modulus was obtained from stress-strain (σ-ε) curves as the slope of initial linear portion corrected by toe compensation. Five samples were tested for each measurement. Measurements of differential scanning calorimetry (DSC) were performed with a Perkin-Elmer Jade DSC covering the temperature range 25-100 °C. The samples were heated at 10 °C/min, and data collected were processed by Perkin-Elmer Pyris 6 software. From the melting peak of the polymer, the crystallinity degree Xc% of the polymer was calculated according to the equation:31 Xc % ¼

ΔHm ° ΔHm

!  100

where ΔHm is the enthalpy of melting measured in the first run and ΔH°m the enthalpy of melting of a totally crystalline PCL (ΔH°m = 139 J/g).32 Peptide Conjugation. PCL-NH2 sheets were bioactivated in a two-step method by using glutaraldehyde (GA) as cross-linking agent to immobilize GRGDY and GYDGR peptides. In the first step, the sheets were treated with a 2% glutaraldehyde (GA) in 10 mM phosphate buffer (PBS) (pH 7.4) at room temperature in an orbital shaker for 3 h. The reaction was stopped by rinsing the samples extensively with deionized water. Afterward, the peptide amino-terminus was covalently linked by reductive amination onto PCL aldehyde activated surfaces. The coupling solutions were made by 50 mM carbonate (pH 8.5) buffer with 0.2 mg/mL of peptides and 5 mM of NaBH3CN. The coupling step was allowed to react for 4 h with gentle shaking and then stopped, rinsing surfaces with copious amount of deionized water. Unreacted aldehyde groups were blocked by treating the polymer surface with 0.2 M ethanolamine in carbonate buffer (50 mM, pH 8.5) for 30 min at room temperature. Finally, the polymercoupled peptide surfaces (PCL-GA-GRGDY, PCL-GAGYDGR) were rinsed with water containing 0.3% Tween 20 and with deionized water; then, samples were stored dry until further use. Afterward, the disks were washed in PBS buffer and sterilized for cell attachment studies or thoroughly washed with (31) Taddei, P.; Simoni, R.; Fini, G. J. Mol. Struct. 2001, 317, 565–566. (32) Crescenzi, V.; Manzini, G.; Calzolari, G.; Borri, C. Eur. Polym. J. 1972, 8, 449–463.

DOI: 10.1021/la100207q

9877

Article

Causa et al.

ultrapure water and dried under vacuum for surface characterization. Analysis of Surface Topology. The surface topography as well as the roughness of all the prepared surfaces was measured with a diCaliber atomic force microscope (Veeco Instruments) in tapping mode in air with a standard silicon tip. During the measurements, the relative room humidity was about 30%, and the room temperature was 25 °C. Images were recorded using height and phase-shift channels with 256  256 measurement points (pixels). Measurements were made several times on different zones of each sample on a scanning area of 90 μm  90 μm, and roughness parameter (arithmetic average Ra, and root-meansquare, rms) calculation and image processing were performed using the SPMLabAnalysis software (Veeco Instruments).

Contact Angle Measurements and Surface Free Energy (SFE) Evaluation. The static contact angle of ultrapure water over the surface of PCL, PCL-NH2, and PCL-GA-GRGDY was measured with automatic video-based measurements of contact angle performed at 25 °C and 65% relative humidity by using an CAM 200 instrument (KSV, Finland). For this measurement, 2 μL of ultrapure water was initially placed over the surface of the polymer. The pictures were processed by KSV-CAM software to calculate the contact angle with the surface polymer by applying the Young/Laplace method. Eight independent measurements were performed per treatment. The statistical significance of the contact angle value was assessed by performing one-way ANOVA with Tukey’s post hoc comparison (INERST v1.3 script in Excel) by comparing both RGD-conjugated and aminated PCL surfaces with respect to the plain polymer. A significance level of 99% (p = 0.001) was chosen for all the tests. Measurements of SFE were performed by evaluating static contact angles of three different liquids (Millipore water, glycerol, and tritolyl phosphate (Aldrich)) onto the different surfaces. At least five measurements were made for each sample and then averaged. The SFE values were evaluated by using the ChaudhuryGood-van Oss model,33 where the total SFE (γTOT) is given by the sum of apolar Lifshitz-van der Waals (γLW) and polar Lewis acid-base (γAB) components: γTOT ¼ γAB þ γLW The contribution of individual acid (γA) and basic (γB) to total polar Lewis acid-base (γAB) were calculated according to the following equation: qffiffiffiffiffiffiffiffiffiffiffiffiffiffi γAB ¼ 2 γA 3 γB

Determination of Conjugated Peptide. The immobilized RGD-peptide was determined directly on solid support by using MicroBCA assay (Sigma-Aldrich) as described from Tyllianakis et al.34 In particular, the number of peptide bonds and the presence of four particular amino acids (cysteine, cystine, tryptophan, and tyrosine) are able to reduce one ion Cu2þ to Cu1þ, which forms a chelate complex with two molecules of bicinchoninic acid (BCA) absorbing at 562 nm.35 This moderate purplecolored complex allows a spectrophotometric determination of nanomolar quantities of functional groups in aqueous solution. This evaluation is finally carried out through standard curves of appropriate substance. The amount of reduction is proportional to peptide bonds present. In our case, the amount of immobilized short peptides, containing the enhancing tyrosine residue (GRGDY and GYDGR), was (33) van Oss, C. J. Colloids Surf., B: Biointerf. 1995, 5, 91–110. (34) Tyllianakis, E. P.; Kakabakos, S. E.; Evangelatos, G. P.; Ithakissios, D. S. Anal. Biochem. 1994, 219, 335–340. (35) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal. Biochem. 1985, 150, 76–85.

9878 DOI: 10.1021/la100207q

performed by adding the MicroBCA working solution directly onto the samples in a reduced volumetric form of the assay (see more in Supporting Information). The determination of the nanomolar peptide concentration was carried out by calibration curve, previously obtained at the same conditions (Figure S2). A nominal density was calculated by taking into account the area of each treated sample and referred to as RGD nmol/cm2. Such a nominal density has to be considered as the overall number of peptide moles immobilized on a polymer chain rather than an effective RGD surface density. In addition, in order to finely investigate the effect of any contaminant on the assay color formation, aminolyzed substrates and pure polymer were also tested. Spatial Distribution of Surface Treatment. Confocal laser scanning microscopy (CLSM) was used to investigate the penetration depth of treatment by LSM 510 Zeiss confocal inverted microscope equipped with a Zeiss 20/3 NA objective and an argon laser (excitation = 541 nm; emission = 572 nm). In order to follow the advance of bioactivation in the two stages of processes (aminolysis and peptide bound polymer), sample surfaces were covalently coupled with two different fluorescent labels. First, PCL-NH2 sheets were treated with 0.1 mg/mL of Rhodamine B isothiocyanate (RBITC) in IPA overnight at 4 °C. Afterward, plates were rinsed thoroughly with IPA and then with 0.3% Tween 20 and water for 24 h to remove any noncovalently bound dye molecule. Conversely, fluoresceinamine (FLUO) was linked to the aldehyde activated polymer surfaces in the same conditions used for peptide conjugation (0.2 mg/mL of dye and 5 mM of NaBH3CN in 50 mM carbonate buffer, pH 8.5). Finally, sheets were removed from reaction wells and copiously rinsed with 0.3% Tween 20 and d.i. water for 24 h to remove any noncovalently linked molecules. The films were then left to dry overnight in a vacuum desiccator before analysis. Nonfunctionalized PCL surfaces were equally processed as control. Samples after Rhodamine B isothiocyanate or fluoresceinamine conjugation were respectively visualized using the characteristic wavelength of Rhodamine (λex = 543 nm; λem = 572 nm) or fluorescein (λex = 496 nm; λem = 518 nm). To compare the results, CLSM settings, in particular laser power, pinhole aperture, detector gain, and amplifier offset, were kept constant for both kinds of observations. Z-stack acquisitions were preformed through the samples by starting from the outer part with optical slices of 1.46 μm. Intensity profiles of fluorescent dyes as a function of penetration depth were obtained along the line drawn in direction perpendicular to the surface. Cell Adhesion Study. Mouse embryo fibroblasts NIH3T3 were maintained at 37 °C and 5% CO2 in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS, BioWhittaker, Walkersville, MD), 2 mM L-glutamine (Sigma, St. Louis, MO), 1000 U/L penicillin (Sigma, St. Louis, MO), and 100 mg/L streptomycin (Sigma, St. Louis, MO). For the experiments, 70-80% confluent cells were used. For cell adhesion experiments, PCL, PCL-NH2, PCL-GAGYDGR, and PCL-GA-GRGDY substrates were sterilized with antibiotics and preincubated in serum-free medium for 16-18 h. After incubation, 5  104 cells were seeded on the samples and grown in DMEM-w/o FBS to avoid unspecific cell adhesion depending on serum protein adsorption. In order to evaluate cell adhesion and shape, scanning electron microscopy (SEM) analysis was performed by Leica 420 by using a beam energy of 5 KeV. After 24 h from cell seeding, the sheets were rinsed with PBS and fixed with 2.5% glutaraldehyde (pH 7.4) (Sigma-Aldrich) for 2 h at room temperature. The cell-plate constructs were dehydrated in graded ethanol concentrations (from 50% to 100% v/v in ethanol), air-dried, gold sputtered, and analyzed by SEM. For CLSM analysis, samples were fixed, after 24 h from cell seeding, with 4% para-formaldehyde for 20 min at RT, rinsed twice with PBS buffer, and incubated with PBS-BSA 0.5% to block unspecific binding. Actin microfilaments were stained with phalloidin tetramethylrhodamine B isothiocyanate Langmuir 2010, 26(12), 9875–9884

Causa et al.

Article

(Sigma-Aldrich). Phalloidin was diluted in PBS-BSA 0.5% and incubated for 30 min at RT. Images were acquired by using a He-Ne excitation laser at the wavelength of 543 nm by using a 20 objective. Moreover, cell viability and proliferation on proposed substrates were evaluated by using Alamar Blue assay (AbD Serotec Ltd., UK) and compared to plain PCL surfaces (see Supporting Information).

Results PCL surfaces were bioactivated by covalently linking RGD motifs directly on surfaces enriched in amino groups after aminolysis (Scheme 1). The aminolysis represents an easy-toperform chemical technique to engraft amino groups along polyester chains, and its limited influence on bulk material properties is often used in surface functionalization of scaffolds for tissue engineering.13,21-23,27,30 In order to study the aminolysis effects onto PCL, a very smooth and nonporous surface was fabricated with a controlled crystallinity degree and surface morphology. PCL pellets were shaped in slight sheets with an average thickness about 200 μm by hot compression molding above the polymer fusion temperature. Such treatment allows reaching Ra and rms average values, respectively, of 0.029 and 0.036 μm without any apparent pore present on the surface. The fusion enthalpy of pure PCL sheets evaluated by DSC was 65.2 ( 5.4 J/g, and the average calculated degree of crystallization was 46.9 ( 3.9%. Chemical modification of PCL sheets was carried out by reaction with a bifunctional amine (DEA) in mild conditions above the amine melting temperature (28 °C). The reaction mechanism was investigated through analysis of surface before and after aminolysis in the proposed reaction conditions. Mechanisms of Aminolysis Reaction and Amine Determination. ATR spectrum of neat PCL showed characteristic peaks of carbonyl and aliphatic groups, respectively, at 1725 cm-1 (ν CO) and 2944, 2865 cm-1 (νs, νas C-H), whereas spectra of aminolized surfaces highlighted more diagnostic peaks related to the presence of amide groups (Figure 1). Peaks more significant are centered at 3338 cm-1 (ν NH), 1641 cm-1 (ν CO), and 1541 cm-1 (δ NH) ascribable to amide I and amide II bands.36 Aminolysis reaction, indeed, takes place by nucleophilic attack at the carbonyl of PCL by diamine. Under alkaline conditions, in an aprotic, polar solvent, the tetrahedral intermediate is deprotonated leading to the formation of an amide and an alcohol.37 To quantify the amine density onto polymer surfaces, a slight modification procedure based on Kaiser test was performed. It is a fast and convenient methodology routinely used to evaluate the amount of peptide growing directly onto resins in solid-phase peptide synthesis.29,38 Differently from others,23,27,30 here the aminolized polymer sheets were brought in solution in order to measure the overall amount of amino groups engrafted onto the chains. By adding first potassium cyanide in pyridine solution and then the other kit components directly onto PCL samples, the reaction between ninhydrin and amino groups takes place in the homogeneous phase. At the final stage of the assay, the purple polymer mixture is suspended in a DCM/ethanol (1:1) mixture, being careful to add first the organic solvent and then ethanol dropwise to avoid polymer precipitation. Dichloromethane resulted in a good solvent for PCL at room temperature and does not affect UV absorption of Ruhemann’s purple

pigment. Evaluation of amino functional groups was carried out through the calibration curve obtained as described in detail in Supporting Information (Figure S1). The nominal density value was obtained by dividing the number of moles obtained for each sample by its surface area (expressed in cm2). It is worth noticing that in the case of a polymeric substrate such as PCL this value has to be referred to as overall amino groups engrafted onto chains rather than areal density. A plot of amine nominal density against time of treatment was obtained at different temperatures (data not shown). The rates of grafting at 24, 37, and 40 °C, respectively, of 1.01, 3.27, and 3.74  10-9 mol/cm2 min were calculated from the initial slopes of the curves between 0 and 30 min. On this basis (see Supporting Information, Figure S3), the activation energy (Ea) of the aminolysis reaction of 69.5 ( 4.6 kJ/mol was evaluated by calculating the linear fitting of log-log (basis e) curves of rates as a function of the temperature.39 This positive value of Ea suggests a strong dependence of the amination as a function of temperature. As already reported in the literature by Croll et al.13 for PLGA and by Bech et al.39 for PET, the Ea is affected by the size of diamine, reported as mobility of molecule, by the solvent, and eventually by the substituent groups flanking the ester. For a linear polyester polymer such as PCL, the chosen aminolysis conditions, such as the 1,6-hexandiamine/isopropanol solution and 37 °C, are a good compromise between the achievement of a fast treatment (rate of grafting of 3.27  10-9 mol/cm2 min) and a temperature low enough to avoid the softening of the samples. A maximum of nominal amino density is obtained after 30 min of treatment (Figure 2A), reaching a value around 100 nmol/cm2. With increasing aminolysis time, the density drops down at a nominal concentration value around 66 nmol/cm2. At the same time, AFM measurements evidenced an increase in roughness parameters in the case of long surface treatment time (longer that 30 min). In particular, Ra and rms were almost constant by increasing the treatment time with a stepwise gain in surface roughness parameters represented by the time of 30 min at 37 °C (Figure 2A). According to the results already reported in the literature, aminolysis has to be referred to as a degradation reaction occurring upon contact with diamine solution. In the first stage, indeed, the reaction starts preferentially at the amorphous regions of the polymer.40 At longer aminolysis time, the decrease of bound NH2 may be caused by chain scission, formation of oligomers and other low mass fragments that are removed

(36) Coates, J. In Encyclopedia of Analytical Chemistry; Meyers, R. A.; Wiley J. & Sons Ltd: Chichester, 2000; pp 10815-10837. (37) Bunnett, J. F.; Davis J. Am. Chem. Soc. 1960, 82, 665–674. (38) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Anal. Biochem. 1970, 34, 595–598.

(39) Bech, L.; Leipottvin, B.; Roger, P. J. Polym. Sci., Part A 2007, 45, 2172– 2183. (40) von Burkersrod, F.; Schedl, L.; G€opferich, A. Biomaterials 2002, 23, 4221– 4231.

Langmuir 2010, 26(12), 9875–9884

Figure 1. ATR-FTIR of plain and aminolyzed PCL substrates. PCL substrate show characteristic extra peaks relative to the presence of amide and amine covalently linked to the polymer surface.

DOI: 10.1021/la100207q

9879

Article

Causa et al.

Figure 2. (a) Plot of nominal amine density and surface roughness parameters (Ra, rms) for aminolyzed PCL substrate for different reaction lengths. Data are reported in terms of mean, and bars represent standard deviation. (b) Residual weight and elastic modulus kinetics during aminolysis treatment (left). A focus on the first 3 h of treatment (right). Data are reported in terms of mean, and bars represent standard deviation.

from the surface during reaction, and the rinsing process.39 The obtained result could thus be due to the scission and detachment of polymer chains present on the surface and the presentation of fresh polymer located beneath. The proposed mechanism could also lead to a plateau value in amine density occurring at longer time of treatment. Influence of Aminolysis Treatment on Bulk Properties. The influence on bulk properties of aminolysis of PCL sheets was investigated through an analysis of degradation kinetics in term of polymer weight loss as well as an evaluation of mechanical and thermal properties during the treatment. A slight decrease (lower than 0.3%) in mass is recorded up to 24 h (Figure 2B left). The rate of degradation is more pronounced in the first 3 h of aminolysis, reaching a mass loss value of about 0.2% after 1 h. A longer time of treatment does not considerably affect the mass. The same degradation kinetics was evaluated by mechanical tests in tensile condition. In particular, elastic modulus variation with treatment time ranging from 15 min to 24 h is reported (inset of Figure 2A right). As result, elastic modulus drops at the early stage of treatment around 3 h; in particular, in the first 2 h the elastic modulus is roughly constant around a value of 300 MPa as reported in the (41) Rosa, D. S.; Guedes, C. G. F.; Bardi, M. A. G. Polym. Test. 2007, 26, 209– 215.

9880 DOI: 10.1021/la100207q

literature in the case of untreated PCL.41 Evaluation of fusion enthalpy during the the aminolysis process did not show any significant change (p = 0.48, n = 8, unpaired t test), highlighting no appreciable modification in the degree of crystallinity. All the experimental results, thus, demonstrate that the proposed aminolysis treatment does not affect the bulk properties of the native PCL. Evaluation of Surface Free Energy and Hydrophilic Characteristics of Treated Surfaces. The change of the polar characteristics of the PCL surface after aminolysis reaction and grafting with RGD-based peptide was investigated in terms of contact angle measurements and of the SFE (with dispersive and polar components) evaluation. The results obtained from contact angle measurements reveal the increase of PCL surface wettability after aminolysis reaction with respect to the untreated PCL surface (Figure 3A) (statistical relevance p < 0.01). The treated surfaces (37 °C, 30 min, DEA 10% w/w) showed an increase in hydrophilic nature of the PCL substrate with a mean contact angle decreasing from about 78° to 60° confirming the incorporation of highly hydrophilic amine groups on the polymer surface. Instead, the subsequent link of the RGD peptide through GA insertion determines only a slight increase of the mean value of the contact angle without any statistically relevant difference with respect to aminated surfaces. The difference between plain and Langmuir 2010, 26(12), 9875–9884

Causa et al.

Figure 3. (a) WCA and SFE evaluation for each step of the procedure reported as mean and standard deviation. (b) Topological mapping of treated surface and spatial distribution of conjugated molecules. Topology (left) observed by AFM in tapping mode of PCL-NH2 surface after 30 min of aminolysis in 10% DEA/IPA solution. Fluorescent microscopy images (right) of PCL after FLUO conjugation.

peptide-conjugated substrates can be ascribable to the presence of the RGD peptide partially neutralizing the charges on the PCLGA-GRGDY sample surfaces. Moreover, while the total SFE (γtot) remains about the same after the aminolysis and RGD peptide treatment the relative weight of the Lifshitz-van der Waals (γLW) and Lewis acid-base (γAB) component changes (Figure 3A). Similar data were reported in the case of adsorption of RGD on PCL surface after plasma treatment.16 In particular, the presence of amine groups and peptide sequences influence acid-base contributions to SFE. This last effect is due to the combined decrease of the acid term γA and the small, but critical, increase of the γB component in terms of the well-know Chaudhury-Good-van Oss equation. RGDfunctionalized surfaces thus exhibited a slightly higher polar character (a higher acid/base contribution) when compared to the untreated surface. This may be attributed to the presence of polar amino acids such as arginine (guanidine group) and aspartic acid (carboxyl group). Evaluation of Grafted Peptide and Morphological Mapping. One-pot colorimetric assay based on the BCA-Cuþ purple color complex was used to determine directly on solid PCL support the amount of immobilized peptides. This assay is widely used to measure proteins both in solution and on adsorbing solid substrates with high sensitivity (picomolar scale) and in a very reproducible way. Furthermore, this method can measure short peptides incorporating cysteine, cystine, tryptophan, and tyrosine that are recognized as cuprous reduction enhancer residues. Yu et al.42 used the BCA-assay to quantify peptides covalently immobilized on hydrogel scaffolds indirectly subtracting the assayed amount of unbound peptide from the known amount of starting (or prebound) peptide. Tyllianakis et al.34 described one incubation step method based on BCA assay as a useful tool (42) Yu, T. T.; Shoichet, M. S. Biomaterials 2005, 26, 1507–1514.

Langmuir 2010, 26(12), 9875–9884

Article

for the determination of solid supports (agarose beads and sepharose gel) functionalized with cysteine and tyrosine. The detection limit of the assay (0.7 nmol/tube for solid tyrosine ligands functional support) was shown as well as the stability of bicinchonic-Cuþ complex and its absorbance in the operative condition and with a series of contaminants (i.e., Tween, SDS, etc.). The quantitation of the different groups was finally carried out through standard curves of an appropriate substance. Here, a MicroBCA assay was used to directly quantify the surface concentration of adhesion peptide covalently bound on PCL substrates by comparison with the calibration curve obtained using the corresponding peptide as standard. The immobilized signal (GRGDY) is an RGD-like sequence containing at the carboxyl end the bulky and lipophilic tyrosine residue that mediates adhesion with high affinity via Rvβ3 and RIIbβ3 integrin receptors1 and increase the sensitivity of the assay.34 The concentration of peptide covalently coupled to the PCL was calculated by taking into account a correction factor obtained by topological analysis. The sample surface area resulted to be about 1.1% higher than the projected area, and then the final result attained a nominal density of 2.81 ( 0.35 nmol/cm2. Surface spatial distribution as well as the penetration depth of treatment was studied by probing the presence of fluorescent moieties. To this purpose, two fluorescent molecules were used as a model to investigate the spatial distribution and penetration depth during each step of the proposed treatment. RBITC was used to label primary amines presented on the PCL substrate, while FLUO was used as a tracer of aldehyde functional groups on which peptide sequences can attach in the last step of conjugation after tether insertion. Confocal microscopy images showed an uneven spatial distribution of FLUO onto PCL surfaces that seems to mimic the amorphous portion of semicrystalline polymer surface (Figure 3B right) reproducing the topology of spherulities at the surface after aminolysis. Such a result suggests a clustering of peptide sequences immobilized on PCL substrate. Moreover, a high depth penetration of treatments was verified. The profile density reported (Figure 4A,B) showed a remarkable penetration that could be ascribable to molecule (FLUO and RBITC) permeability within the polymer. In particular, a cross section of functionalized samples showed a fluorescence emission down to about 60 μm from the upper surface in the case of aminolysis for RBITC conjugation, while the penetration was reduced to 50 μm in the case of FLUO conjugation with a significant reduction of intensity at about 30 μm. Such results emphasize the overestimation of evaluated densities for both amine and peptide. Bioactivation Assessment Through NIH3T3 Fibroblast Adhesion Tests. In order to study the bioactivation of PCL substrates at the cell-material interface, their interaction with fibroblast cells was studied. SEM images revealed that, after 24 h from seeding, NIH3T3 cells adhere on all the substrates. However, cell morphology drastically changes on different samples. In particular, on nontreated PCL, PCL-NH2, and PCL-GAGYDGR surfaces, cells showed a round shape, indicating a scarce adhesion to the substrate (respectively Figure 5 A-E, B-F, and C-G). Conversely, on PCL bioactivated with RGD peptide, cells were correctly adhered and well-spread on the surface, showing a good interaction with the material (Figure 5D,H). Moreover, at higher magnification, the formation of filopodia was observed (Figure 5I). The effect of PCL functionalization in enhancing cell adhesion was further confirmed by actin cytoskeleton staining. This qualitative analysis indicated that cells were better adhered on RGD bioactivated substrates (Figure 5M), as demonstrated by the presence of stress fibers, compared to cells seeded on DOI: 10.1021/la100207q

9881

Article

Causa et al.

Figure 4. Intensity profiles of fluorescent molecules in the depth during the two-step conjugation procedure: (a) CLSM intensity profile as a function of depth in the z-direction of PCL-NH-RBITC. (b) CLSM intensity profile as a function of depth in the z-direction of PCL-GAFLUO samples.

Figure 5. Scanning electron microscope micrographs (A and E PCL; B and F PCL-NH2; C and G PCL-GA-GYDGR; D, H, and I PCL-GAGRGDY). Confocal laser scanning microscope images of phalloidin staining of microfilaments (J PCL; K PCL-NH2; L PCL-GA-GYDGR; and M PCL-GA-GRGDY). Bar 50 μm.

nontreated PCL, PCL-NH2, and PCL-GA-GYDGR surfaces (Figure 5J,K,L), where there is no evidence of cytoskeleton organization. Moreover, Alamar blue assay (see Figure 4 in Supporting Information) confirmed a significant improvement in cell proliferation of RGD-conjugated substrate after two days of culture when compared to plain PCL. 9882 DOI: 10.1021/la100207q

Discussion The binding of amino acid sequences encoded along extracellular protein backbones (such as fibronectin, collagen, vitronectin, and laminin) to cell membrane integrins is one of the key mechanisms by which these cells recognize the material.1 The most studied peptide-functional materials are those that mimic Langmuir 2010, 26(12), 9875–9884

Causa et al.

the integrin binding RGD sequences. More importantly, the discussion about the importance of spatial distribution of the presenting peptide signal, its density, and tether choice in influencing cell response is still open as well as chemistry with respect to roughness contribution to cell adhesion.43 Chemical methods and, in particular, surface chemical modification rather than chemi- or physisorption represents a robust, repeatable, and easy to perform procedure to bind peptide sequence on polymeric surfaces. However, to realize bioactive interfaces it is mandatory to investigate the surface characteristics as well as the presentation of immobilized signals with respect to the polymeric substrate. Here, we monitored the evolution of chemical and physical properties of polymeric surface during each step of the bioconjugation. In particular, the synthesis consisted of the aminolysis of the polyester substrate to obtain functional groups on the surface, on which peptide sequences were coupled by a homobifunctional cross-linker in a second step. Aminolysis is already reported as a procedure to introduce primary amines on polyesters such as PLGA,13 PET, or PCL.27 Data already reported in the literature from Gabriel et al.22 as well as Zhu et al.27 showed, in different treatment conditions, lower or almost equal amine density on PCL: respectively, 7.9  10-9 and 2  10-7 mol/cm2, while an increase of roughness after aminolysis was also evidenced on PCL23,27 and on PET fibers by Roger et al.39 In particular, as already described in the literature,27 the maximum NH2 density yielded at 30 min is about 10-7 mol/cm2, that in the case of flat surfaces and all amino groups laid as single layer should correspond to an average area per amino-terminated chain of about 0.1 A˚2. Such a result represents an impossibly high concentration. Indeed, even though surface roughness is not taken into account, data presented in this work attaining around 60 nmol/cm2 should be referred to as a surface layer of about 60 μm (see Figure 4A) and, thus, corrected on the basis of density profile along the depth of the sample. Taking into account penetration of aminolysis treatment and, in particular, supposing a constant level of amine in the first 60 μm of PCL sheets is possible to estimate the extent of primary modification equivalent to about 0.16% by considering polymer density of 1.12 g/cm3. As reported by Croll et al.13 for PLGA substrates, a level of primary modification below 1% is theoretically required to allow the attachment of a complete monolayer of proteins via covalent binding. In our case, a coverage of about 10% should be achieved on the PCL surface. Differently, our estimation should be more realistic by considering the amount of amine distributed in a micrometric thickness rather than distributed merely on the surface. It was also reported that the aminolysis resulted in the formation of an amide bond and an alcoholic group on the PCL backbone with the following lysis of chain segments undergoing functionalization.27 Such a phenomenon could bring about a plateau value after a peak in the number of amines present on PCL due to the detachment of the interested portion of the PCL chain at a longer time of treatment. Crack formation as well as a massive loss of mechanical properties after partial degradation during aminolysis is already reported, mainly in the case of PET.44,45 The influence on bulk properties and the control over the surface chemistry and evolution during the treatment of its physicochemical properties were studied as key aspects. In particular, it was demonstrated that the bulk mechanical properties are not significantly affected by aminolysis in the proposed (43) Perlin, L.; MacNeil, S.; Rimmer, S. Soft Matter 2008, 4, 2331–2349. (44) Haghighat Kish, M.; Borhani, S. J. Appl. Polym. Sci. 2000, 78, 1923–1931. (45) Holmes, S. A. J. Appl. Polym. Sci. 1996, 61, 255–260.

Langmuir 2010, 26(12), 9875–9884

Article

reaction conditions. A negligible decrease of elastic modulus (∼11%) and a very slight reduction of mass (lower than 0.3%) were obtained after 3 h of treatment. However, it worth noticing that aminolysis of PCL also presents some limitations concerning the range of engrafted amine groups, falling in the range from a few tens to one hundred nmol/cm2, as well as the lack of control of their spatial distribution, impairing a uniform distribution of signals on the polymeric surface because of its semicrystalline nature. As for the bioactivation after RGD conjugation, there is a general consensus for solid and rigid materials, such as glass, assuming that the amount of peptide is evenly distributed over the surface of the substrates. It is generally considered that only a modification of surface is achieved and that the reagent penetration depth is the same as the integrin accessible depth of 10 nm, treating results in a two-dimensional manner.46 In the case of polymers, instead, the depth of the surface modification after treatment will depend upon polymer crystallinity, size of pore, and swelling capability of polymer at surface.43 Here, we found a high penetration depth that could determine a large overestimation of peptide density that can contribute to cell adhesion. A corrected density can be obtained by calculating the penetration depth of conjugated fluorescent dye (assumed as peptide molecule) into the polymer bulk immobilized at the same conditions used for the GRGDY conjugation. In detail, the density of bioactive GRGDY peptide covalently immobilized through a tether represented by a short aldehyde were first obtained by the MicroBCA method; it was assumed that the only GRGDY molecules available for the cytoskeletal-associated transmembrane receptors, represented by integrins, were located in an outer layer of 10 nm of polymer.46 Such a distance can be considered a threshold distance for the integrin engagement; afterward, the amounts of molar of GRGDY obtained by MicroBCA were corrected on the basis of the profile density obtained along the depth (fluorescent intensity profile was obtained by FLUO as a model molecule). A corrected surface density of about 1 pmol/cm2 was estimated as the number of molecules per area available for the integrin engagement. The relevance of the quantity of peptide per area can be compared with literature reported showing that a minimum of RGD density of 1.0  10-15 mol/cm2 is sufficient in the case of fibroblast to promote cell spreading, while only a density of 1.0  10-14 mol/cm2 is sufficient to promote the formation of focal contact in the case of glass substrates.4 Equally, more recent publication reported a density of 10-12 mol/cm2 of conjugated RGD peptide on PET surfaces as sufficient for an improvement in cell adhesion.47 However, it is worth emphasizing that such results are related to the particular surface characteristics and chemistry; that is, in our case, a higher roughness and different presentation of RGD-based peptide to cell integrins could determine different thresholds levels. Moreover, it is widely reported in the literature that, beyond ligand density, the morphology and spatial organization of the presenting solid signals at the surface and, in particular, of the bioactive peptide strongly affect cell behavior and fate. Recently, it was also demonstrated that integrin clustering and cell adhesion induced by RGD ligands is dependent on ligand arrangement and, in particular, that the cell adhesion is inhibited for a RGD nanopattern order while activated by a nanopattern disorder if ligand spacing is higher than 70 nm.48 On this basis, the (46) Brandley, B. K.; Schnaar, R. L. Anal. Biochem. 1988, 172, 270–278. (47) Chollet, C.; Chanseau, C.; Remy, M.; Guignandon, A.; Bareille, R.; Labrugere, C.; Bordenave, L.; Durrieu, M. C. Biomaterials 2009, 30, 711–720. (48) Huang, J.; Gr€ater, S. V.; Corbellini, F.; Rinck, S.; Bock, E.; Kemkemer, R.; Kessler, H.; Ding, J.; Spatz, P. J. Nano Lett. 2009, 9, 1111–1116.

DOI: 10.1021/la100207q

9883

Article

Causa et al.

Conclusions

organization of the bioactive ligands immobilized on the polymer surface was studied through topological investigation and fluorescence mapping. A fluorescent molecule (FLUO) was selected to mimic the peptide in studying its distribution on the PCL surface. The presence of uneven distribution of fluorescent signal on treated surfaces demonstrated a non-uniform molecular conjugation at the micrometric level (Figure 3B right), while a disordered topography can be recognized at the nanometric level (Figure 3B left). The lack of spatial uniformity in peptide conjugation can be ascribable to the intermediate step of aminolysis, resulting in enhancement of the amorphous portion of the polymer surface.49 This motivates a similar appearance at the surface of the topology of aminated PCL (Figure 3B left) with the fluorescence images of FLUO-conjugated PCL (Figure 3B right). Furthermore, the overall gain in hydrophilic characteristics was already demonstrated to improve cell adhesion on polymeric surface.50,51 In our case, the same presence of a peptide sequence can alter the physical and chemical properties of the background materials and affect the biological properties of the materials through nonspecific means. The control, represented by scrambled peptide sequence, confirms the specific binding of immobilized peptides. Moreover, adhesion tests were carried out in serum-free conditions. The effect of serum absorption at the cell-material interface can, indeed, potentially interfere with the peptide engagement by integrin receptors. Even though far from in vivo conditions, the serum free condition was chosen to rule out any effect of serum proteins in the adhesion mechanisms, being more appropriate when investigating the presentation of bioactive molecule to cell integrins. The enhancement in cell adhesion is thus fully ascribable to the effective engagement of RGD sequence by integrin receptors (presumably Rvβ3 and RIIbβ31) present on NIH3T3 cell membrane. Moreover, viability tests confirmed that the conjugated peptides significantly improve the NIH3T3 cell proliferation on PCL substrate, indicating that the proposed conjugation of GRGDY peptide can play a crucial role in building up biomimetic materials, especially for tissue engineering applications.

Acknowledgment. The authors thank Dr. Antonio Gloria for mechanical tests, Dr. Gobind Das for helping us in spectroscopic measurements, and Dr. Maurizio Ventre for statistical analysis.

(49) Zeronian, S. H.; Collins, M. J. Text Prog. 1989, 20, 1. (50) Lim, J. Y.; Shaughnessy, M. C.; Zhou, Z.; Noh, H.; Vogler, E. A.; Donahue, H. J. Biomaterials 2008, 29, 1776–1784. (51) Arima, Y.; Iwata, H. Biomaterials 2007, 28, 3074–3082.

Supporting Information Available: Additional information and figures as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

9884 DOI: 10.1021/la100207q

It is becoming increasingly apparent that the future of synthetic biomaterials will depend on the ability to develop materials to trigger a specific function in cell behavior at the cell-material interface. To do this, a fundamental step is represented by the investigation of effective presentation of ligands toward cell component and the control of surface characteristics after biomaterial treatment. In cell-material interaction, cell-adhesion represents a fundamental cell function to be regulated. To this regard, we investigated the immobilization of RGD peptide on PCL surfaces by monitoring the surface characteristics after each step of treatment and evaluating ligand distribution on the surface and penetration in the polymer substrate. Here, we demonstrated that aminolysis represents an easy route to introduce primary amines with high yield that can be easily optimized with respect to processing conditions. More importantly, conjugation of amine-terminated peptides by means of reductive amination after tether insertion showed a specific recognition of the solid signal to NIH3T3 integrin cell receptors highlighting a correct presentation of the peptide sequences. In particular, GRGDY conjugation onto PCL revealed a typical spatial distribution that traces the boundary of polymer crystallite at the surface. Moreover, the investigation on treatment penetration revealed that the thickness of immobilized ligand spans tens of micrometers reducing the “effective” signal immobilized on the surface but not affecting bulk mechanical properties of the polymer. A determination of amines and peptides effectively engrafted on polymer was carried out and discussed. These approaches provide a useful procedure to investigate polymeric surfaces after modification with specific molecules and highlight the potential for realizing and characterizing surfaces capable of improving cell attachment or, in the case of different peptide sequences or structure, promote specific cell-material interactions.

Langmuir 2010, 26(12), 9875–9884