ARTICLE IN PRESS

Biomaterials 27 (2006) 691–701 www.elsevier.com/locate/biomaterials

Fabrication, characterization, and biological assessment of multilayered DNA-coatings for biomaterial purposes

Jeroen J.J.P. van den Beuckena, Matthijn R.J. Vosb, Peter C. Thu¨ nec, Tohru Hayakawad, Tadao Fukushimae, Yoshio Okahataf, X. Frank Walboomersa, Nico A.J.M. Sommerdijkb, Roeland J.M. Nolteb,g, John A. Jansena,Ã

aDepartment of Periodontology and Biomaterials, College of Dental Science, Radboud University Nijmegen Medical Center, P.O. Box 9101, 6500 HB Nijmegen, the Netherlands bLaboratory for Macromolecular and Organic Chemistry, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, the Netherlands cSchuit Institute of Catalysis, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, the Netherlands dDepartment of Dental Materials, Research Institute of Oral Science, Nihon University School of Dentistry at Matsudo, 2-870-1 Sakaecho Nishi, Matsudo, Chiba 271-8587, Japan eDepartment of Dental Engineering, Bioengineering section, Fukuoka Dental College, 2-15-1 Tamaru, Sawara-ku, Fukuoka 814-0193, Japan fDepartment of Biomolecular Engineering, Tokyo Institute of Technology, 4259 Nagatsuda, Midori-ku, Yokohama, Kanagawa 226-8501, Japan gDepartment of Organic Chemistry, Radboud University Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, the Netherlands

Received 30 May 2005; accepted 21 June 2005 Available online 1 August 2005

Abstract

This study describes the fabrication of two types of multilayered coatings onto titanium by electrostatic self-assembly (ESA), using deoxyribosenucleic acid (DNA) as the anionic polyelectrolyte and poly-D-lysine (PDL) or poly(allylamine hydrochloride) (PAH) as the cationic polyelectrolyte. Both coatings were characterized using UV-vis spectrophotometry, atomic force microscopy (AFM), X-ray photospectroscopy (XPS), contact angle measurements, Fourier transform infrared spectroscopy (FTIR), and for the amount of DNA immobilized. The mutagenicity of the constituents of the coatings was assessed. Titanium substrates with or without multilayered DNA-coatings were used in cell culture experiments to study cell proliferation, viability, and morphology. Results of UV-vis spectrophotometry, AFM, and contact angle measurements clearly indicated the progressive build-up of the multilayered coatings. Furthermore, AFM and XPS data showed a more uniform build-up and morphology of [PDL/DNA]- coatings compared to [PAH/DNA]-coatings. DNA-immobilization into both coatings was linear, and approximated 3 mg/cm2 into each double-layer. The surface morphology of both types of multilayered DNA-coatings showed elevations in the nanoscale range. No mutagenic effects of DNA, PDL, or PAH were detected, and cell viability and morphology were not affected by the presence of either type of multilayered DNA-coating. Still, the results of the proliferation assay revealed an increased proliferation of primary rat dermal fibroblasts on both types of multilayered DNA-coatings compared to non-coated controls. The biocompatibility and functionalization of the coatings produced here, will be assessed in subsequent cell culture and animal-implantation studies. r 2005 Elsevier Ltd. All rights reserved.

Keywords: AFM; Cell culture; Cell morphology; Cell proliferation; Cell viability; Electrostatic self-assembly; Fibroblast; FTIR; MIT assay; Mutagenicity; Nanotopography; SEM; Surface modification; Titanium

1. Introduction

In implantology, a wide variety of materials are used ÃCorresponding author. Tel.: +31 24 3614006; fax: +31 24 3614657. to generate biomedical devices with satisfactory proper- E-mail address: [email protected] (J.A. Jansen). ties. Although bulk properties of materials largely

0142-9612/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2005.06.015 ARTICLE IN PRESS 692 J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 determine their suitability for a given application, the the PEM structure. Instead, changes in the ionic content biological response is mainly determined at the bioma- of the polyelectrolyte solutions in the fabrication process terial surface, via the interactions with components of are one of the means to modulate PEM properties, the biological surroundings [1,2]. including layer thickness [14,24]. Modifications of the biomaterial surface have been a Irrespective of the progress made so far, the use of major research topic in the past decades. In view of this, DNA-containing multilayered coatings in the biomater- an intriguing novel coating material for biomaterials is ials field still requires their complete characterization. deoxyribonucleic acid (DNA), the repository of (in- Whereas the build-up mechanism of PEMs has been dividual-specific) genetic information. Irrespective of its described previously, information on the morphological genetic information, the structural properties of DNA properties is scarce. Additionally, the hypothesized give this unique, natural compound potential for use as beneficial properties of DNA necessitate the quantifica- a biomaterial coating. The specific build-up of the DNA tion of the amount of DNA immobilized in the coatings. molecule may ensure a versatile use at various implanta- Therefore, the aim of this study was to fabricate and tion sites. The molecular structure of DNA in vertebrate characterize two types of multilayered DNA-coatings, species is homogenous [3,4], and the non- or low- which differ in the type of cationic polyelectrolyte used, immunogenic properties of DNA (compared to other and assess their effects on cell behavior using in vitro- biological antigens like proteins and sugars) may limit experiments. both innate and acquired immune responses [3,5,6]. Additionally, DNA can incorporate other molecules via groove-binding and intercalation [7,8]. This creates opportunities to specifically deliver desired biological 2. Materials and Methods mediators in the direct vicinity of the implantation site. 2.1. Materials and fabrication Finally, the high phosphate content in DNA might, via the high affinity of phosphate for calcium ions [9,10], 2.1.1. Materials beneficially affect the deposition of calcium in the bone Polyanionic DNA (7300 bp/molecule; sodium salt) was formation process. The use of DNA as a functional kindly provided by Nichiro Corporation (Yokosuka-shi, biomaterial, instead as a carrier for its genetic informa- Kanagawa prefecture, Japan). Potential protein impurities in tion, has already been suggested [11,12], and pioneering the DNA were checked using the BCA protein assay efforts have resulted in the fabrication of DNA-contain- (Pierce, Rockford, Illinois, USA) and measured to be below ing bulk (bio)material [13], which demonstrated to cause 0.20% w/w (data not shown). Polycationic polyelectrolytes no adverse reactions upon subcutaneous implantation in poly(allylamine hydrochloride) (PAH; MW 70,000) and the backs of rats. poly-D-lysine (PDL; MW 30,000–70,000) were purchased from The application of DNA as a coating material, Sigma (Sigma-Aldrich Chemie B.V., Zwijndrecht, the Nether- however, is hampered by (a) its easy nucleolytic lands). All materials were used without further purification. The monomeric structures of the cationic polyelectrolytes used degradation, and (b) its solubility in aqueous solutions. are presented in Fig. 1. The use of the previously mentioned DNA-containing bulk material [13] resulted in easily detaching coatings on various substrates. Therefore other methods to 2.1.2. Substrate preparation and cleaning obtain stable DNA-containing coatings for biomaterial purposes have to be explored. One applicable technique Three types of substrates were used: could be electrostatic self-assembly (ESA), also known as layer by layer (LbL) assembly. Via this technique, 1. Quartz substrates (45 12 mm; Hellma Benelux B.V., polyelectrolyte multilayers (PEMs) can be generated Rijswijk, the Netherlands) were used for UV-vis spectro- through the alternated progressive adsorption of oppo- photometry, contact angle measurements, FTIR spectro- sitely charged polyelectrolytes via electrostatic interac- scopy, and XPS. tions [14]. 2. Silicon substrates (1 1 cm; Wafernet GmbH, Eching, The ESA technique has been applied successfully Germany) were coated with a 50 nm thick titanium layer, using DNA as a polyanionic building block [15–22]. using the RF magnetron sputter technique. The titanium 4 Interestingly, electrostatic interactions with positively sputtering process (duration 5 min, pressure 5 10 charged polyelectrolytes have demonstrated to protect mbar) was performed with the silicon substrates attached to a rotating sample holder. Titanium-sputtered silicon DNA from degradation by nucleolytic activity [23]. substrates were used for XPS, contact angle measurements, Furthermore, the method of fabrication (i.e. under and AFM. immersion in aqueous solutions) inherently evidences 3. Disc-shaped titanium substrates (12 mm diameter, 2 mm the water insolubility of PEMs. Even immersion of thickness; commercially pure titanium, as-machined) were PEMs in high ionic solutions (i.e. exceeding those in used to determine the incorporation of radiolabeled DNA physiological conditions) does not cause dissociation of into each double-layer. ARTICLE IN PRESS J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 693

2.2. Characterization

2.2.1. UV-Vis spectrophotometry The progressive build-up of the multilayered DNA-coatings onto quartz substrates was monitored using a Shimadzu n multispec UV-Vis spectrophotometer (Shimadzu Germany, Duisburg, Germany), equipped with a diode-array detector. Measurements were taken after each successive layer. Growth of the multilayered DNA-coatings was studied via the + - incremental increase at 260 nm, the absorbance maximum (A) NH3 Cl of the nucleic base chromophores. All measurements were replicated in a separate second run. H NH - C -CO OH n CH2 2.2.2. Atomic Force Microscopy (AFM) CH 2 The morphology of non-coated titanium-sputtered silicon, CH2 partial ([PDL/DNA]½ and [PAH/DNA]½; [PDL/DNA]2 and [PAH/DNA]2; [PDL/DNA]3½ and [PAH/DNA]3½) and CH2 complete ([PDL/DNA]5 and [PAH/DNA]5) multilayered (B) NH2 DNA-coatings was analyzed using a Nanoscope IIIa AFM setup (Digital Instruments, Buffalo, NY, USA). Fig. 1. Molecular structure of (A) poly(allylamine hydrochloride) All scans were made in contact mode, using a Si3N4-tip. (PAH) and (B) poly-D-lysine (PDL). Typical scan sizes were between 0.5 0.5 mm2 and 5 5 mm2. To detect potential damaging effects of the probe to the multilayered DNA-coatings, scans of the samples were always made with increasing scan size. Image analysis Quartz substrates were cleaned in Piranha solution (H2O2 software (WSxM free software, downloadable at http:// (30% aqueous solution)/H2SO4—3:7 v/v) to completely remove all traces of organic materials at the surface, and www.nanotec.es) was used to generate micrographs and increase the hydrophilicity of the surface (caution: Piranha quantitatively compare surface roughness, using root- solution is an extremely corrosive mixture). Subsequently, the mean-square (RMS) calculations, a roughness denotation substrates were thoroughly rinsed with ultra-pure water from a based on the standard deviations of the z-value (vertical Millipore system (resistivity X18.2 MO cm; Millipore B.V., direction). Amsterdam, the Netherlands) and dried using a pressurized air stream. 2.2.3. X-ray photospectroscopy (XPS) Titanium-sputtered silicon substrates and titanium discs XPS measurements were performed using a non-mono- were cleaned ultrasonically using consecutively nitric acid chromized VG-Escalab 200 spectrometer equipped with an (10% v/v), acetone, and isopropanol. Between each cleaning aluminum anode (Al KR; 1486.6 eV) operating at 510 VA step, the substrates were rinsed with ultra-pure water. After the with a background pressure of 2 109 mbar. Spectra were last cleaning step, the substrates were dried in air. acquired at 01 and 601 with respect to the surface normal, after the adsorption of 1, 3½, and 5 double-layers of [PDL/DNA] 2.1.3. Generation of multilayered DNA-coatings or [PAH/DNA] onto titanium-sputtered silicon substrates. As Multilayered DNA-coatings were generated using the ESA a reference, the spectrum of a dropcast (pure) DNA film was technique, as described by Luo et al. [17] with few modifica- included. The (IP2p/IN1s) (ratio of integrated intensities) of the tions. The cleaned substrates were immersed in an aqueous experimental substrates were used to calculate the fraction of solution of either PDL (0.1 mg/ml) or PAH (1 mg/ml) for DNA (xDNA) within the top (5–10 nm) of the measured 30 min, allowing sufficient time for the adsorption of the first sample, according to Formula 1: cationic polyelectrolyte layer onto the substrates. Subse- ÀÁ I =I ¼ x I =I . (1) quently, the substrates were washed in ultra-pure water P2p N1s DNA P2p N1sDNA (5 min, continuous water flow) and dried using a pressurized As the stoichiometric carbon/nitrogen ratios for DNA, PDL, air stream. Thereafter, the substrates were alternately im- and PAH are similar, no extra correction was necessary. mersed in an anionic aqueous DNA solution (1 mg/ml) and the respective cationic polyelectrolyte solution for 7 min each, with intermediate washing in ultra-pure water (5 min, continuous 2.2.4. Contact angle measurements water flow) and drying using a pressurized air stream. The The progressive build-up of the multilayered DNA-coatings build-up of the multilayered DNA-coatings was continued on quartz and titanium-sputtered silicon substrates was until a total of 5 double-layers was reached. These were monitored using contact angle measurements on a Drop designated either [PDL/DNA]5 or [PAH/DNA]5) (the denota- Shape Analysis System DSA10 (Kru¨ ss GmbH, Hamburg, tion of coatings is restricted to indicating the number of Germany). Four measurements per substrate were taken using 1 double-layers, i.e. 2 represents only the cationic part of the the sessile drop method. All measurements were replicated in a double-layer). separate second run. ARTICLE IN PRESS 694 J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701

2.2.5. Fourier transform infrared spectroscopy (FTIR) of the 4th or 5th culture passage. Prior to the assays, cells were Infrared spectra of the multilayered DNA-coatings on detached using trypsin/EDTA (0.25% w/v trypsin/0.02% quartz substrates were obtained using Fourier transform EDTA) and concentrated by centrifugation at 1500 rpm for infrared spectroscopy (Perkin Elmer Instruments, Bucks, 5 min. Subsequently, cells were resuspended in culture medium England, UK). The infrared spectrum of DNA (as sodium and counted using a Coulters counter (Beckman Coulter Inc., salt) was obtained using the KBr-pellet method (0.4% w/w Fullerton, CA, USA). For all in vitro assays, cells were seeded DNA in KBr). at 1 104 cells/cm2 into 24-well plates (Greiner Bio-One BV, Alphen aan de Rijn, the Netherlands). 2.2.6. Build-up of multilayered DNA-coatings with radiolabeled DNA 2.3.3. Cell proliferation The amount of DNA immobilized into the multilayered At 1, 3, 7, and 10 days after seeding, the medium was coatings was determined using radiolabeled DNA. In brief, aspirated and the cell layer was washed twice with PBS. DNA was radiolabeled using 32P-postlabeling through the Subsequently, substrates with attached cells were transferred T4-kinase enzyme. Non-bound label was removed via 2 into fresh wells and incubated with 0.5 ml trypsin/EDTA. successive filtrations using Sephadexs-50 columns. The After detachment of the cells, 0.5 ml culture medium was efficiency of labeling was 34%. Radiolabeled DNA was added added. The cell suspension was diluted using Coulters Isotone to an aqueous DNA solution with a final concentration of II diluent (Beckman Coulter Inc.) and cell numbers were 1 mg/ml. Concentrations of the aqueous solutions of the counted using a Coulters counter. The experiment was cationic polyelectrolytes were as described above. Multilayered performed 3 times, with at each time point 3 samples per coatings were fabricated as described above onto disc-shaped condition ðn ¼ 3Þ. glass and titanium substrates. After the completion of 1–5 double-layers, substrates were taken out of the fabrication 2.3.4. Cell viability process, immersed in scintillation fluid, and counted using a Cell viability was assessed using a commercially available liquid scintillation counter. For comparison, samples contain- cytotoxicity assay, according to the instructions of the ing 100 ml of the initial aqueous DNA solution (1 mg/ml) were manufacturer (Promega Corporation, Madison, WI, USA). counted. All experimental and control samples were present in The test is based on the conversion of a tetrazolium salt ([3-(4,5- 3-fold. dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide], MTT) into a formazan product. Briefly, at 3 days after cell seeding 2.3. Biological assessment on the experimental substrates, 150 ml dye solution was added to each well, and cells were incubated at 37 1Cfor4h. 2.3.1. Mutagenicity assay Subsequently, 1 ml of solubilization/stop solution was added The mutagenicity of the chemicals used to build up the to each well, and the cells were incubated for another hour (at multilayered DNA-coatings was assessed using the Muta- room temperature). Finally, the contents of the wells were ChromoplateTM basic kit (version 3.1; Environmental Biode- mixed thoroughly using a pipette and the absorbance at 570 nm tection Products Inc., Brampton, On., Canada), according to wavelength was read using a spectrophotometer. Measured the instructions of the manufacturer. This kit is based on the absorbance values were corrected for cell numbers using parallel most generally used and validated bacterial reverse-mutation samples, from which the cell number was determined using a s test, known as the ‘Ames Test’. Briefly, a mutant strain of Coulter counter. Normalization of the data was performed to Salmonella typhimurium (TA100), carrying a mutation in the the results of the non-coated control substrates. operon coding for histidine biosynthesis, was exposed to 2 concentrations of the chemicals used for the build up of the 2.3.5. Cell morphology multilayered DNA-coatings (DNA and PAH: 250 and To study cell morphology, cells cultured on the experi- 2500 mg/ml; PDL: 25 and 250 mg/ml). After an incubation mental substrates for 3 days were washed twice with PBS. period of 5 days at 37 1C, the number of positive wells for each Subsequently, cell fixation was carried out for 15 min. in 2% condition was recorded and scored against the results of the glutaraldehyde in 0.1 M sodium cacodylate buffered solution. background (negative control). Sodium azide (NaN3;5mg/ml), Then, samples were rinsed twice with cacodylate buffered a direct-acting mutagen, was used as a positive control. solution and dehydrated using a graded series of ethanol. Finally, samples were dried using with tetramethylsilane. The 2.3.2. Cell culture and seeding samples were sputter coated with gold, and examined using a Biological experiments were performed using DNA-coatings JEOL 6310 SEM. generated on disc-shaped titanium substrates and non-coated titanium control substrates. All substrates were sterilized using 2.3.6. Statistical analysis a UV-irradiation treatment (254 nm; 4 h). Statistical analyses were performed using GraphPad Instat, Rat dermal fibroblasts (RDF) were obtained from the version 3.0 (GraphPad Software, San Diego, CA, USA). ventral skin of male Wistar rats, using a standard procedure as Statistical comparisons for RMS-values within one type of described previously [25], after which cells were cryo-pre- coating were performed using an ANOVA-test, combined served. Before the initiation of an experiment, cells were with a post hoc Tukey–Kramer Multiple Comparisons Test. thawed and cultured in culture medium (a-MEM (Gibco), Statistical comparisons between the two types of coatings were supplemented with 10% v/v fetal calf serum (FCS) and performed using an unpaired t-test. The results of the gentamicin (50 mg/ml)). All assays were performed with cells biological assays were analyzed using an unpaired t-test. ARTICLE IN PRESS J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 695

3. Results 0.3

3.1. Characterization 0.2 3.1.1. UV-Vis photospectrometry The build-up of multilayered DNA-coatings on 0.1 quartz substrates via UV-Vis spectrophotometry could absorption (a.u.) be easily monitored using the nucleic base chromo- phores in DNA (absorbance maximum at 260 nm). As 0 illustrated in Fig. 2 A and B, both multilayered DNA- 0123456 coatings showed an increase in UV-absorbance spec- (A) Number of PDL-DNA double layers trum with every successively added double-layer. The 0.1 UV-absorbance of [PDL/DNA] double-layers was high- er compared to that of [PAH/DNA] double-layers. 0.08 When peak values at 260 nm were plotted against the number of double-layers (Figs. 3A and B), a linear 0.06 increase with every added double-layer became apparent for both types of coatings (Figs. 3A and B). 0.04 absorption (a.u.)

0.02 3.1.2. Atomic Force Microscopy 0123456 The surface morphology of partial and complete (B) Number of PAH-DNA double-layers multilayered DNA-coatings on titanium-sputtered sili- Fig. 3. Plots of the absorbance at 260 nm (A260) against the number of con substrates was studied using AFM. The forces of the (A) [PDL/DNA]-double-layers, and (B) [PAH/DNA]-double-layers. cantilever on the coatings during contact mode scanning were well below the threshold causing damage to the Fig. 4 shows the AFM height images of (partial and coatings, as evidenced by the absence of damage on complete) multilayered DNA-coatings of both [PDL/ AFM-images with increasing scan size. DNA] (Fig. 4A–E) and [PAH/DNA] (Fig. 4F–J) double- layers. For both types of coating, the AFM height images showed an optical increase in surface roughness. Further analysis of the surface roughness using RMS- values of both types of multilayered DNA-coatings showed significant increases with every step (i.e. 1½ double-layer) from [PDL/DNA]1/2 and [PAH/DNA]2 during the build-up of the coatings (Table 1). Further- more, RMS-values of [PDL/DNA]-coatings were sig- nificantly lower than those of equivalent [PAH/DNA]- coatings after the deposition of 2 double-layers. Typical differences in surface morphology between the two types of coatings included the spatial distribu- tion of elevations, as well as their average height. Whereas [PDL/DNA]5 showed a relatively homogenous morphological appearance with equally spaced eleva- tions of relatively low height (average 6 nm), [PAH/ DNA]5 showed randomly distributed elevations of relatively large height (average 14 nm). To further illustrate these differences, 3D-image reconstructions of both complete multilayered DNA-coatings are presented in Fig. 5.

3.1.3. X-ray photospectroscopy (XPS) The XPS spectra of DNA (dropcast) and 1, 3½, and 5 double-layers of [PDL/DNA] or [PAH/DNA] on Fig. 2. UV-absorption spectra of (A) [PDL/DNA]-multilayer coat- ings, and (B) [PAH/DNA]-multilayer coatings. The UV-absorption titanium-sputtered silicon substrates are presented in spectra were monitored using quartz substrates after the deposition of Fig. 6. The XPS spectrum of DNA (dropcast) showed 0 (bl), 1 (a), 2 (b), 3 (c), 4 (d), and 5 (e) double-layers. the presence of magnesium (Mg), probably as a cationic ARTICLE IN PRESS 696 J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701

Fig. 4. AFM height images of several stages during the build-up of a [PDL/DNA]5 (A–E) and [PAH/DNA]5 (F–J) multilayered coating on titanium- sputtered silicon substrates. Images were taken after the adsorption of 0 (A and F), ½ (B and G), 2 (C and H), 3½ (Dand I), and 5 (E and J) double- layers.

Table 1 Determination of surface roughness (RMS (in nm)7standard devia- tion)

(Partial) coating [PDL/DNA] [PAH/DNA]

Control (titanium-sputtered silicon) 1.6470.03 1.6470.03

[double-layer]½ 1.6870.27 1.9970.10 Ã [double-layer]2 2.4070.11 1.9670.11 Ã Ã [double-layer]3½ 3.1670.17 4.5870.96 Ã Ã [double-layer]5 5.4970.57 8.3271.70

ÃAsterisk indicates significantly increased surface roughness (po0:05) compared to previous (partial) coating.

Fig. 6. XPS spectra of dropcast DNA film and indicated number of double-layers for (A) [PDL/DNA] or (B) [PAH/DNA]. Scaling of the spectra is adjusted for constant integrated intensity of Phosphorus 2p emissions (not shown).

counter ion, whereas almost no sodium (Na) or potassium (K) peaks could be detected. Correspond- ingly, XPS spectra of DNA-terminated [PDL/DNA]- coatings showed the presence of Mg-peaks. However, DNA-terminated [PAH/DNA]-coatings did not show the presence of Mg-peaks, nor did coatings terminated with a cationic layer (PDL or PAH). Furthermore, a distinct difference was observed regarding the presence of titanium (Ti) peaks. Whereas partial and complete [PAH/DNA]-coatings showed the presence of Ti-peaks, equivalent [PDL/DNA]-coatings did not. Fig. 5. Three-dimensional images of the morphology of a [PDL/ No angle dependency was observed regarding the

DNA]5 and [PAH/DNA]5 multilayered coating on titanium-sputtered phosphorus/nitrogen ratio (IP2p/IN1s) for the (dropcast) silicon substrates. DNA film, as expected from a one-component film. ARTICLE IN PRESS J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 697

Interestingly, (partial) [PDL/DNA]- and [PAH/DNA]- coatings also showed no angle dependency, implying a homogeneous distribution of DNA in the surface region of the coatings. The amount of DNA in the top (5–10 nm) of the 3 multilayered coatings was calculated using Formula 1 6 DNA 963 9 5 8 3 3 8 2 (see Section 2). Table 2 demonstrates that [PDL/DNA]- 6 0 1235 1 0 1088 1 coatings display a higher DNA content in the DNA- 1063 1 terminated layers compared to [PAH/DNA]-coatings. %T 8

3.1.4. Contact angle measurements 8

0 [PDL/DNA]5 Regardless of the substrate used (either quartz or 1088 1 titanium-sputtered silicon), the contact angles demon- 3 1

strated to be higher after the adsorption of the positively 2 charged polymer (either PDL or PAH) than after the 1213 1

adsorption of the negatively charged DNA (data not 8

8 [PAH/DNA] shown). When the build-up of both types of multi- 0 5 1088 1 layered DNA-coatings progressed, the alternating be- 4 2 havior of the contact angles became more pronounced. 2 1224 1 1000 450 3.1.5. Fourier transform infrared spectroscopy (FTIR) Wavelength (cm-1) DNA exhibited its characteristic bands of base-pairs 1 1 1 at 1602 cm (adenine), 1650 cm (thymine), 1684 cm Fig. 7. FTIR-spectra of DNA (upper line), [PDL/DNA]5 (middle (guanine), and 1481 cm1 (cytosine) [26]. For character- line), and [PAH/DNA]5 (bottom line). Both types of multilayered DNA-coatings were build-up on quartz substrates. ization, however, spectra with minimum wavelengths of 1250 cm1 were used, since PDL and PAH do not have characteristic infrared spectral criterions below this wavelength. 80 Infrared spectral criterions of the DNA double helical structure can be observed in the 1250–950 cm1 sugar- 60 phosphate backbone vibration range (Fig. 7; upper line). 40 DNA showed its characteristic absorption bands at 1235, 1088, 1063, and 963 cm1, which can be assigned 20 to the antisymmetric stretching of the phosphate groups 1 0

(nas 1235 cm ) [27], the symmetric stretching vibrations DNA content (ug/substrate) 1 2 3 4 5 1 2 3 4 5 1 of the phosphate groups (ns 1088 and 1063 cm ) [18,19], [PDL/DNA] [PAH/DNA] and the CC stretching of the backbone (963 cm1) [27], Fig. 8. Immobilization of DNA into each double-layer during the respectively. build-up of [PDL/DNA]5 and [PAH/DNA]5 coatings on titanium Both the [PDL/DNA]5 (Fig. 7, middle line) and substrates. Measurements were performed using radiolabeled DNA. [PAH/DNA]5 (Fig. 7, lower line) multilayered coatings showed the characteristic infrared absorption band at 1088 cm1, corresponding with the n band. Further- s 3.1.6. DNA immobilization during the build-up of more, a frequency shift of the nas absorption band is 1 multilayered coatings visible for [PDL/DNA]5 (1235-1213 cm ) and for [PAH/DNA] (1235-1224 cm1). The amount of DNA (mg/substrate) immobilized into 5 the multilayered DNA-coatings was analyzed using radiolabeled DNA. In Fig. 8, the progressive immobi- Table 2 lization of DNA into multilayered DNA-coatings Percentage of DNA in top (5–10 nm) of the coating on the titanium is represented. Results of the DNA-immobili- experimental substrates (calculations using Formula 1) zation into multilayered DNA-coatings on glass sub- strates were also performed (data not shown). Whereas Coating architecture % DNA in [PDL/ % DNA in [PAH/ DNA] DNA] on glass substrates the amount of DNA immobilized in the first double-layer is higher for [PDL/DNA]-coatings [double-layer]1 87 52 than for [PAH/DNA]-coatings (33 vs. 20 mg/substrate, [double-layer]3½ 66 59 respectively), an equal amount of DNA (20 mg/ [double-layer] 80 72 5 substrate) is immobilized into the first double-layer of ARTICLE IN PRESS 698 J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 both types of multilayered DNA-coatings on titanium ated using scanning electron microscopy. Fig. 10 substrates. After the generation of the first double-layer, presents representative images of these cells after a both types of multilayered DNA-coatings immobilize a 3-day culture period. Although the native roughness of constant amount of DNA with each successively the as-machined titanium substrates does not allow a adsorbed double-layer, irrespective of the substrate clear observation of completely spread cells, it becomes material (i.e. glass or titanium) onto which the multi- apparent from slightly elevated cells that both types of layers are adsorbed. The amount of DNA immobilized multilayered DNA-coatings do not alter the morpholo- into the multilayered DNA-coatings is approximately gical appearance of the cells compared to non-coated 9–10 mg DNA per double-layer (E3 mg/cm2). titanium controls. In general, all cells had a typical fibroblast appearance with several (short) cellular 3.2. Biological assessment extensions.

3.2.1. Mutagenicity assay Incubation of a mutant strain of Salmonella typhi- 4. Discussion murium (TA100) in selected concentrations of DNA (250 and 2500 mg/ml), PDL (25 and 250 mg/ml), or PAH The aim of this study was to fabricate and character- (250 and 2500 mg/ml) did not show a significant increase ize multilayered DNA-coatings, generated via the ESA- in reverse-mutation (p40.05; data not shown). Negative technique. The choice of DNA as the anionic building and positive controls showed a limited and an intensive block relates to the structural properties of the DNA- increase in reverse-mutation, respectively. molecule and their potential beneficial effects in implantology rather than to the encoded genetic 3.2.2. Cell proliferation and viability information. As the cationic building blocks, either The results of a representative cell proliferation assay PDL or PAH was used. are presented in Fig. 9. Compared to non-coated control Many researchers utilize a polyelectrolyte, which under neutral acidity is extremely charged (e.g. poly(ethylenei- substrates, significantly higher cell numbers (po0:05) were observed at almost every assayed time point on mine), PEI; or poly(styrenesulfonate), PSS) as the both types of multilayered DNA-coatings. No differ- primary layer(s) onto the substrate of desire ences in cell proliferation between the two types of [15,19,20,22,28–32]. This is done to achieve optimal multilayered DNA-coatings were observed. coverage of the oppositely charged substrate. However, To assess potential effects of multilayered DNA- other researchers do not indicate this necessity coatings on cell viability, the mitochondrial redox [17,33–36]. To avoid potential detrimental effects of activity was assessed using a MTT-based assay. Neither PEI on cellular behavior [28], and because the build-up of the two types of multilayered DNA-coatings sig- of PEMs is well-acceptable without using extremely nificantly decreased mitochondrial redox activity of charged polyelectrolytes as a primary layer, the use of RDF cells after 3 days of culture (data not shown). polyelectrolytes for our two types of multilayered DNA- coatings was limited to the functional anionic polyelec- trolyte DNA, and one specific cationic polyelectrolyte 3.2.3. Cell morphology (either PDL or PAH). The morphological appearance of the primary fibro- The build-up of multilayered DNA-coatings has been blast cells, cultured on both types of multilayered DNA- demonstrated to be dependent on the type of cationic coatings and non-coated titanium controls, was evalu- polyelectrolyte used. Linear build-up regimes have been demonstrated for PAH-containing coatings [14,21,37], whereas exponential build-up regimes were found for 250000 PLL-containing coatings [36,38]. Although the driving control [PDL/DNA] [PAH/DNA] * * 200000 force for both growth regimes is essentially similar (i.e. the charge overcompensation with each deposited 150000 polyionic layer), the diffusive behavior of PLL through- 100000 * * out the coatings allows an exponential build-up, * proportional to the amount of PLL that diffuses out 50000 ** of the coating during immersion in a polyanionic 0 solution [38,39]. However, based on both the UV- Cell counts (cells/substrate) 17103 spectrophotometry results and the results of the immobilization of radiolabeled DNA in our study, we Fig. 9. Proliferation of RDF cells on non-coated titanium controls, conclude that the build-up regime for both our [PDL/ [PDL/DNA]5-, and [PAH/DNA]5-coated titanium substrates. Results are shown as mean7SD. Asterisks indicate significant difference DNA]5 and [PAH/DNA]5 coatings is linear, with typical compared to non-coated control (po0:05). incremental increases with each successively adsorbed ARTICLE IN PRESS J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 699

the first double-layer was found for multilayered DNA- coatings. Probably, these differences relate to the amount of cationic polyelectrolyte adsorbed in the primary adsorption layer and the inherent specific affinity of the used cationic polyelectrolytes for each type of substrate. The structure of the DNA after immobilization into either the [PDL/DNA]5 or [PAH/DNA]5 coating re- mains unclear. The shift of the nas band from 1235 to 1 1213 cm for [PDL/DNA]5 coatings and from 1235 to 1224 cm1 for [PAH/DNA]-coatings may reflect a change in the conformation from A to Z-form and from A to B-form, respectively [27]. Shifts of the nas band after complexation of DNA with cationic polyelectrolytes or (A) ions have been observed frequently [18,19,40], and are explained by the interaction of the anionic phosphate groups of the DNA-molecules with cationic counter- parts. Additionally, the relative humidity of multilayered DNA-coatings is an important factor determining the precise location of characteristic bands [19]. Never- theless, the results of the FTIR spectroscopy clearly indicate the presence of double-helical DNA within both types of multilayered DNA-coating, as evidenced by the presence of the characteristic absorption band at 1088 cm1 [19]. The biological relevance of DNA structure in the coatings remains to be determined. AFM images of both types of multilayered DNA- coatings clearly demonstrate a nanoscale roughness for both [PDL/DNA]5 and [PAH/DNA]5. The difference in RMS-roughness calculation between equal numbers (B) of [PDL/DNA]- and [PAH/DNA]-double-layers most probably also relates to the adsorption of the first cationic polyelectrolyte layer. The more homogenous [PDL/DNA]5-coating (RMS ¼ 5.49 nm) reflects a uni- form adsorption of the primary PDL-layer, whereas for the [PAH/DNA]5-coating (RMS ¼ 8.32 nm) randomly distributed elevations suggest a heterogeneous adsorp- tion of the primary PAH-layer. This is further evidenced by the results of the contact angle measurements, which show a more pronounced alternating behavior after the adsorption of several double-layers. XPS-data corroborate with the results of the AFM- measurements. The build-up of [PDL/DNA]-coatings is more homogeneous than that of [PAH/DNA]-coatings, reflected by the presence of Ti-peaks and the absence of Mg-peaks in the latter coatings. A less homogeneous (C) DNA incorporation and/or an irregular build-up regime of the multilayer structure for [PAH/DNA]-coatings Fig. 10. SEM images showing the morphology of RDF cells on (A) might explain these results. Additionally, the absence of non-coated titanium controls, (B) [PDL/DNA]5-, and (C) [PAH/ Mg-peaks in the DNA-terminated [PAH/DNA]-coat- DNA]5-coated titanium substrates after 3 days of cell culture. Note the presence of multiple cellular extensions (arrowheads) of the fibroblasts ings might be an indication that in this case all counter on all types of substrates. ions of the DNA are exchanged by the positively charged PAH-component, resulting in a decreased double-layer of approximately 10 mg DNA on titanium 3Dstructure of the DNA-molecules. Hence, Ti-peaks substrates, respectively. Interestingly, a substrate-depen- might be present in [PAH/DNA]-coatings due to the dent difference in the amount of DNA immobilized in limited thickness of these coatings. The absence of ARTICLE IN PRESS 700 J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701

P angle-dependent differences in the /N-ratios for both Medical Center, Nijmegen) and Dr. Gerald Verhaegh types of coatings indicates a homogeneous distribution for their help with the AFM analyses and the of the molecules in the top (5–10 nm) of the coatings. radiolabeled DNA-experiments, respectively. AFM Therefore, the layers can be regarded as not completely imaging was performed in collaboration with the separated, but partly mixed. This mixed character of Microscopical Imaging Center of the Nijmegen Center PEMs has been described previously [14]. for Molecular Life Sciences (NCMLS). Mr. Maichel van The mutagenicity assay demonstrated that the mate- Riel (Control Laboratory, Central Animal Facility, rials used to generate both types of multilayered DNA- Radboud University Nijmegen Medical Centre) is coatings are non-mutagenic. Further biological assess- acknowledged for his assistance with performing the ment revealed that fibroblast cell proliferation was mutagenicity assay. This study was financially sup- increased on both types of multilayered DNA-coatings ported by the Dutch Technology Foundation STW, compared to non-coated titanium controls. Although grant # NKG.5758. the reason remains unclear, the increased surface roughness [41–44] or an altered protein adsorption profile on both types of multilayered DNA-coatings might affect fibroblast cell proliferation. Nevertheless, References we attribute the elevated cell growth to the presence of a multilayered DNA-coating, and not for instance to [1] Kasemo B, Gold J. Implant surfaces and interface processes. Adv Dent Res 1999;13:8–20. DNA release from the layers. In view of the studies [2] Castner DG, Ratner BD. Biomedical surface science: foundations performed by Kitamura et al. [45], who demonstrated to frontiers. Surf Sci 2002;500:28–60. that the addition of DNA to the culture medium [3] Krieg AM. Immune effects and mechanisms of action of CpG significantly decreases the proliferation of fibroblast motifs. Vaccine 2000;19(6):18–22. cells, this furthermore indicates the stability of the [4] Stryer L. Biochemistry, 1995. [5] Krieg AM, Yi AK, Matson S, Waldschmidt TJ, multilayered DNA-coatings. The metabolic activity of Bishop GA, Teasdale R, Koretzky GA, Klinman DM. CpG the fibroblast cells cultured on both types of multi- motifs in bacterial DNA trigger direct B-cell activation. Nature layered DNA-coatings was not affected, indicating that 1995;374:546–9. both types of multilayered DNA-coatings are cytocom- [6] McMichael AJ. Antigens and MHC systems. In: McGee JOD, patible. Isaacson PG, Wright NA, editors. Oxford textbook of pathology. Oxford, UK: Oxford University Press; 1992. In conclusion, this study demonstrates the feasibility [7] Werner MH, Gronenborn AM, Clore GM. Intercalation, of the ESA-technique to fabricate multilayered DNA- DNA kinking, and the control of transcription. Science 1996; coatings. UV-vis spectrophotometry, AFM, and contact 271:778–84. angle measurements clearly indicated the progressive [8] Wilson WD. Reversible interactions of nucleic acids with small build-up of both [PDL/DNA]- and [PAH/DNA]-multi- molecules. In: Blackburn GM, Gait MJ, editors. Nucleic acids in chemistry and biology. Oxford, UK: Oxford University Press; layers. Furthermore, AFM and XPS data showed a 1996. more uniform build-up and morphology of [PDL/ [9] Tretinnikov ON, Kato K, Ikada Y. In vitro hydroxyapatite DNA]-coatings compared to [PAH/DNA]-coatings. deposition onto a film surface-grated with organophosphate The DNA-immobilization into both types of multi- polymer. J Biomed Mater Res 1994;28:1365–73. layered DNA-coatings was linear, and approximated [10] Kamei S, Tomita N, Tamai S, Kato K, Ikada Y. Histologic and 2 mechanical evaluation for bone bonding of polymer surfaces 3 mg/cm into each double-layer, irrespective of cationic grafted with a phosphate-containing polymer. J Biomed Mater polyelectrolyte and substrate material. The surface Res 1997;37:384–93. morphology of both types of multilayered DNA-coat- [11] Yamada M, Kato K, Nomizu M, Sakairi N, Ohkawa K, ings showed elevations in the nanoscale range. Biologi- Yamamoto H, Nishi N. Preparation and characterization of cal assessment demonstrated an increased proliferation DNA films induced by UV irradiation. Chemistry 2002;8: 1407–12. of primary fibroblasts on both types of multilayered [12] Inoue Y, Fukushima T, Hayakawa T, Takeuchi H, Kaminishi H, DNA-coatings compared to non-coated titanium con- Miyazaki K, Okahata Y. Antibacterial characteristics of newly trols. Cell viability and morphology were unaffected by developed amphiphilic lipids and DNA-lipid complexes against the presence of either type of multilayered DNA- bacteria. J Biomed Mater Res 2003;65A:203–8. coating. Further biocompatibility and functionalization [13] Fukushima T, Inoue Y, Hayakawa T, Taniguchi K, Miyazaki K, Okahata Y. Preparation of and tissue response to DNA-lipid of the coatings produced here, will be assessed in films. J Dent Res 2001;80:1772–6. subsequent cell culture and animal-implantation studies. [14] Decher G. Fuzzy Nanoassemblies: toward Layered Polymeric Multicomposites. Science 1997;277:1232. [15] Pei R, Cui X, Yang X, Wang E. Assembly of alternating Acknowledgements polycation and DNA multilayer films by electrostatic layer-by- layer adsorption. Biomacromolecules 2001;2:463–8. [16] Sastry M, Rao M, Ganesh KN. Electrostatic assembly of We thank Dr. Bas Feddes (Department Peridodon- nanoparticles and biomacromolecules. Acc Chem Res 2002; tology and Biomaterials, Radboud University Nijmegen 35:847–55. ARTICLE IN PRESS J.J.J.P. van den Beucken et al. / Biomaterials 27 (2006) 691–701 701

[17] Luo L, Liu J, Wang Z, Yang X, Dong S, Wang E. Fabrication of [32] Tedeschi C, Mohwald H, Kirstein S. Polarity of layer-by-layer layer-by-layer deposited multilayer films containing DNA and its deposited polyelectrolyte films as determined by pyrene fluores- interaction with methyl green. Biophys Chem 2001;94:11–22. cence. J Am Chem Soc 2001;123:954–60. [18] Zhou Y, Li Y. Studies of interaction between poly(allylamine [33] Richert L, Boulmedais F, Lavalle P, Mutterer J, Ferreux E, hydrochloride) and double helix DNA by spectral methods. Decher G, Schaaf P, Voegel JC, Picart C. Improvement of Biophys Chem 2004;107:273–81. stability and cell adhesion properties of polyelectrolyte multilayer [19] Sukhorukov GB, Montrel MM, Petrov AI, Shabarchina LI, films by chemical cross-linking. Biomacromolecules 2004;5: Sukhorukov BI. Multilayer films containing immobilized nucleic 284–94. acids. Their structure and possibilities in biosensor applications. [34] Richert L, Lavalle P, Vautier D, Senger B, Stoltz JF, Schaaf P, Biosens Bioelectron 1996;11:913–22. Voegel JC, Picart C. Cell interactions with polyelectrolyte [20] Sukhorukov GB, Mohwald H, Decher G, Lvov YM. Assembly of multilayer films. Biomacromolecules 2002;3:1170–8. polyelectrolyte multilayer films by consecutively alternating [35] Shi X, Sanedrin RJ, Zhou F. Structural characterization of adsorption of polynucleotides and polycations. Thin Solid Films multilayered DNA and polylysine composite films: influence 1996;220:284–5. of ionic strength of DNA solutions on the extent of DNA [21] Lvov Y, Decher G, Sukhorukov G. Assembly of thin films by incorporation. J Phys Chem B 2002;106:1173–80. means of successive deposition of alternate layers of DNA and [36] Picart C, Lavalle Ph, Hubert P, Cuisinier FJG, Decher G, Schaaf Poly(allylamine). Macromolecules 1993;26:5396–9. P, Voegel J- C. Buildup mechanism for poly(L-lysine)/Hyaluronic [22] Chen X, Lang J, Liu M. Layer-by-layer assembly of DNA-dye acid films onto a solid surface. Langmuir 2001;17:7414–24. complex films. Thin Solid Films 2002;409:227–32. [37] Decher G, Hong JD, Schmitt J. Buildup of ultrathin multilayer [23] Hill IR, Garnett MC, Bignotti F, Davis SS. Determination of films by a self-assembly process: III. Consecutively alternating protection from serum nuclease activity by DNA-polyelectrolyte adsorption of anionic and cationic polyelectrolytes on charged complexes using an electrophoretic method. Anal Biochem 2001; surfaces. Thin Solid Films 1992;210–211:831–5. 291:62–8. [38] Picart C, Mutterer J, Richert L, Luo Y, Prestwich GD, Schaaf P, [24] Schonhoff M. Layered polyelectrolyte complexes: physics of Voegel JC, Lavalle P. Molecular basis for the explanation of the formation and molecular properties. J Phys: Condens Matter exponential growth of polyelectrolyte multilayers. Proc Natl Acad 2003;15:R1781–808. Sci USA 2002;99:12531–5. [25] Freshney RI. Culture of animal cells: a manual of basic technique. [39] Mendez Garza J, Schaaf P, Muller S, Ball V, Stoltz J- F, Voegel New York, USA: Alan R. Liss Inc; 1987. J- C, Lavalle P. Multicompartment Films Made of Alternate [26] Liu W, Sun S, Cao Z, Zhang, X, Yao K, Lu WW, Luk KDK. An Polyelectrolyte Multilayers of Exponential and Linear Growth. investigation on the physicochemical properties of chitosan/DNA Langmuir 2004;20:7298–302. polyelectrolyte complexes. Biomaterials. [40] Zhou Y, Li Y. Layer-by-layer self-assembly of multilayer films [27] Banyay M, Sarkar M, Graslund A. A library of IR bands of containing DNA and Eu3+: their characteristics and interactions nucleic acids in solution. Biophys Chem 2003;104(2):477–88. with small molecules. Langmuir 2004;20:7208–14. [28] Tryoen-Toth P, Vautier D, Haikel Y, Voegel JC, Schaaf P, [41] Flemming RG, Murphy CJ, Abrams GA, Goodman SL, Nealey Chluba J, Ogier J. Viability, adhesion, and bone phenotype of PF. Effects of synthetic micro- and nano-structured surfaces on osteoblast-like cells on polyelectrolyte multilayer films. J Biomed cell behavior. Biomaterials 1999;20:573–88. Mater Res 2002;60:657–67. [42] Thapa A, Miller DC, Webster TJ, Haberstroh KM. Nano- [29] Serrano MC, Pagani R, Vallet-Regi M, Pena J, Ramila A, structured polymers enhance bladder smooth muscle cell function. Izquierdo I, Portoles MT. In vitro biocompatibility assessment of Biomaterials 2003;24:2915–26. poly(epsilon-caprolactone) films using L929 mouse fibroblasts. [43] Zhang R, Ma PX. Synthetic nano-fibrillar extracellular matrices Biomaterials 2004;25:5603–11. with predesigned macroporous architectures. J Biomed Mater Res [30] Zhu H, Ji J, Shen J. Construction of multilayer coating onto 2000;52:430–8. poly (DL-lactide) to promote cytocompatibility. Biomaterials [44] Teixeira AI, Abrams GA, Bertics PJ, Murphy CJ, Nealey PF. 2004;25:109–17. Epithelial contact guidance on well-defined micro- and nanos- [31] Boura C, Menu P, Payan E, Picart C, Voegel JC, Muller S, Stoltz tructured substrates. J Cell Sci 2003;116:1881–92. JF. Endothelial cells grown on thin polyelectrolyte mutlilayered [45] Kitamura H, Takemoto N, Mizuno M, Kuboki Y, Sakairi N, films: an evaluation of a new versatile surface modification. Nishi N. Suppression for the proliferation of fibroblasts by Biomaterials 2003;24:3521–30. external DNA. Int J Biol Macromol 1997;21:337–40.